In this episode

• How to decide free vs paid vs freemium vs trial based on your business

• Why free users can give bad product signal (and how to find the paying 20%)

• The $1 pricing move: why $1→$5 is easier than free→paid

• How PLG can “sell the service layer”

Chapters:

01:00 Free vs paid
02:40 When free makes sense
05:00 Why “broad free” often produces no monetization
06:00 Using free users for QA vs getting misleading feedback
07:00 The “freeness” problem
08:00 The third path
09:05 How to find the paying segment
12:00 How many responses you need
14:00 Example teardown
18:00 Customer research
19:05 Pricing signals
20:35 Why raising prices is easier than going from free to paid
21:45 Should you start at $1/month?
22:45 Freemium vs fully paid
26:55 Rolling out paid upgrades
29:00 Avoiding product drift
32:00 B2B/enterprise differences
37:15 AI pricing in 2026
40:40 Where to find Joe

Resources & links

- Joe Wilkinson: Artisan Growth Strategies
- Joe on X/Twitter: @artisangrowth

If you want to apply this

1. Audit your current model: are you truly building the free engine or funding paid acquisition?

2. Talk to 5 users: “What happened before/after using the product?”

3. Add a net-new paid SKU and try to get a real pre-purchase.

Subscribe / follow!

If you want more episodes on pricing, PLG, and growth strategy for founders, follow the show so new episodes land in your feed.

Also, I’m actively looking to chat with anyone who runs a Slack channel. If you know someone, please hit reply and let me know. 🙏

Hope everybody’s doing good.

/ Ben

Finding product-market fit for hardware with Apple Vision Pro designer

Is hardware is the new SaaS? Who knows, but to get ahead of it - this interview shares practical lessons for indie makers. Key takeaways: cycle 3–5 prototypes (not 50), use Wizard-of-Oz testing to fake features, watch for gift/secondary use cases, lean on influencers + GIFs instead of specs, and mix pre-orders/grants/crowdfunding for early capital.

✍️ Product-Market Fit for Physical Products, (12 minutes)

The trap of tight positioning (as a creator)

Jakob Greenfeld went the build-in-public + personal brand playbook and explains why boxing yourself content-wise can kill momentum. Instead, he suggests how to approach creating more freely.

✍️ Don’t Put Yourself in a Box, by Jakob Greenfeld (5 minutes)

Twitter’s new product head: How to grow a lvl 99 account

Nikita Bier, now Head of Product at Twitter, shared a roadmap for platform growth. Worth a look if you want insight into how the platform might evolve. (Or, you don’t believe it and want this week’s ab workout).

✍️ Nikita Bier on Twitter Growth (3 minutes)

Justin Welsh’s content system in 15 minutes

An oldie but a goodie. This breakdown of Justin Welsh’s content operating system (explained for Kajabi) is in my opinion the free TLDR of his course. Clear and still relevant for anyone trying to systemise their content production.

📺 Justin Welsh Content System Explained (15 minutes)

Why Facebook ad accounts get banned (and how to avoid it)

A friend recently had their ad account banned :-( Let’s avoid this. Here’s a reminder of what to make sure you’re not doing - and what to have in place - so you don’t get canned by the Meta cops.

✍️ Why Facebook Ad Accounts Get Banned (and How to Avoid It), by Benjamin Boman (5 minutes)

P.S. Want to build something together?

Here are a few ways we can work together:

• Validate your product idea: Use HypeDeck.io to get early feedback before building anything. Great if you want feedback on your ideas, not just pre-sales.

• Learn to DIY Facebook ads: I made a course for my friends who asked for a speedrun on DIY lead ads. It’s how I do what I do, condensed into a digital playbook. You can get it too on Gumroad.

• Work with me on your Ads: If you want help running ads or want to bounce around strategy ideas, reach out here.

/ Ben

Why you shouldn’t build a personal brand?

Alex Becker is not for the soft-stomached, but hold on for this one. He gives a counter-perspective on why you shouldn’t focus on building a personal brand - and the better alternative. This might give some confidence and direction if you don’t want to build a brand yourself.

📺 Why Having A "Personal Brand" Will Ruin Your Life, Alex Becker (22 minutes)

How long should your YouTube videos be? Does it matter?

When my mate David told me that video length is a quiet signal about quality, I didn’t fully register the idea- but the more I thought about it, the more it makes sense. This guide breaks down what the optimal length actually is for value-signaling and retention.

📺 I Found Out Which Video Length Makes It SPREAD, Faster (TimeZone Method), Nate Black (20 minutes)

Building a more authentic personal brand by focusing on the work

A mate and I were talking about whether he should build a brand around himself as the niche (all his natural interests) or pick a topic. It reminded me to remind you (ha) that I wrote a post sharing my latest view on this. TLDR - perhaps it’s better to focus on making your work famous (which makes you famous) rather than focusing on building a following first.

✍️ Ryan Reynolds and the Maker-Creator, by me (5 minutes)

GummySearch founder Folio Fed’s early indiehacker journey - Key lessons learned

Fed is an indiehacker who actually made it and this post catches him earlier in the journey, before that. It’s interesting because you can see the perspectives that worked out in hindsight (hint: it didn’t involve focusing 100% on one product).

✍️ Bootstrapping 2 products while consulting and traveling, by Fed (8 minutes)

Make it easier - work to your strengths by Managing oneself (book summary)

This book came up in discussion with some friends recently. It can sometimes feel like you’re smashing your head against the wall when maybe you’re just not working with yourself, but against yourself. This summary gives insights on how to align with your natural proclivities so you don’t have to read the whole book. Definitely brain superfood for thought.

✍️ Managing Oneself, summary by Rick Lindquist (6 minutes)

P.S. Want to build something together?

Here are a few ways we could collaborate:

• Validate your product idea: Use HypeDeck.io to get early feedback before building anything. Great if you're still testing the waters.

• Learn to DIY Facebook ads: I made a course for my friends who asked for a speedrun on DIY lead ads. It’s how I do what I do, condensed into a digital playbook. You can get it too on Gumroad.

• Work with me on your Ads: If you want help running ads or want to bounce around strategy ideas, reach out here.

/ Ben

How to think long-term when the end goal feels impossibly far

Tim Urban’s classic on Musk breaks down a mindset you can actually use: work from first principles, backcast from the end state, and run small “ingredient-gathering” projects until the path appears.

✍️ The Cook and the Chef: Musk’s Secret Sauce, by Tim Urban (21 minutes)

What to post on LinkedIn (that isn’t cringe thought-leadership)

I’ve started playing around with LinkedIn lately (thanks for the love if you’ve seen it) and people always ask what to post beyond “inspirational takes.” This lays out the 3 types I’ve been using.

📺 A 3 Types Of Content You Should Be Posting On LinkedIn, Natasha Vilaseca (9 minutes)

Design a YouTube channel that actually grows and monetises

Great breakdown of channel strategy. Has a clear take on the 4 creator archetypes and how each monetises, then the 6 parts of running the channel.

📺 Welcome to YouTube 2025: Let Me Show You How to Win (22 minutes)

How I structure Facebook lead-gen ad accounts (so they stop breaking)

When I review Meta accounts for friends, this is the #1 thing I wish they knew earlier. Here’s how I suggest structuring an account overall for proper creative testing and also scalability.

✍️ How I structure lead-gen Facebook ad accounts (4 minutes)

How to start a blog like it’s 2009, but smarter

I found this old gem recently in my saved articles list on Pocket (now defunct). Andrew Chen’s OG blogging advice has some unique takeaways that have held up to time.

✍️ How to Start a Professional Blog: 10 Tips for New Bloggers, by Andrew Chen (6 minutes)

How to turn short-form wins into long-form assets

Nicolas Cole and Dickie Bush are some of the best known ‘writers’ on Twitter/Linkedin with over 600,000 followers on X. This is their method for expanding a viral short piece of content into a long form winner - and why.

✍️ Expanding Short-Form into Long-Form Content: 5 Easy Steps, by Nicolas Cole & Dickie Bush (5 minutes)

P.S. Want to build something together?

Here are a few ways we could collaborate:

• Validate your product idea: Use HypeDeck.io to get early feedback before building anything. Great if you're still testing the waters.

• Learn to DIY Facebook ads: I made a course for my friends who asked for a speedrun on DIY lead ads. It’s how I do what I do, condensed into a digital playbook. You can get it too on Gumroad.

• Work with me on your Ads: If you want help running ads or want to bounce around strategy ideas, reach out here.

/ Ben

A lot of people are jumping on the personal brand bandwagon (including me). But something that always felt a bit hollow to me was its most common approach:

Building an audience first, before having a product.

It sometimes leads folks to vapid social media. Performative posting, annoying platitudes, follow for follow, and other distasteful behaviour on social media.

I’m not the only one who thinks this, as Jakob Greenfield suggests it can contribute to the Bullshit Creative Industrial Complex (🤣), where people try to build credibility through curated insight (like this newsletter?) rather than value production. People should Build a business, not an audience.

The other problem is that this sometimes makes it harder to start. What do you post about if you’re not working on something? People have often asked me, a non-expert btw, things like:

• How do I pick my niche?

• What do I post about?

• etc

By chance, I came across an essay written by Darin Persinger on A Peek Inside Ryan Reynolds’s Business Model, on how Ryan used his comedic media skills to co-found a marketing agency that he then uses on his own projects.

Looking at this, I think his story offers some cool ideas on HOW the maker-creator model can be done tastefully, as an opposite to the audience-first approach.

✋ Side note: What’s a maker creator? When I say maker here, I mean makers in the broad sense of the word. People who create things that produce value for others, books, apps, websites, physical products, businesses, etc. And a creator as somebody who creates content, often involving themselves, to promote things.

Lessons for Maker Creators

None of these are rocket science, but a nice encapsulation of the ideas in practice:

1. Avoid personal content brands to improve severability. It’s a silly example, but had Ryan made Reynolds Telecom or Reynolds Gin, I’d argue it’d be much harder for him to have sold out of his stakes along the way. Instead, each project simply featured him in the content, boosting the project’s visibility and naturally building his following that he could direct towards other projects later, too. This is great news for people who have multiple interests, and aligns with the perspective shared by in her article don’t build businesses, build projects.

2. Focus on interesting projects to attract natural followers. People follow Ryan not because they love his lifestyle content or memes on his Instagram (Dagobert’s experience), but because the like his work and then become interested in him. It’s clear from his timeline is that each project grew his popularity. And I’m not suggesting it’s always a business, but I think making it a project of some kind fits this idea. The focus is on making things people love.

3. Use yourself in your project and it’s content. I doubt much of this would have worked had he never appeared in any movies or adverts etc. He’s clearly IN his work, which allows people to discover and want to follow him. An interesting example of this is who puts his name directly on his work as a SaaS founder. (See example).

4. Use your profile for commentary & direct attention. He seems to leverage his social media profiles to talk about his work across different interests, rather than becoming the main and only outlet for content about those things. Like building in public. This is also different to creator brands who lead with their name, only to get locked into a niche later. This came up twice in Nathan Barry’s article, The Billion Dollar Creator, where one creator even added an edit to suggest he wouldn’t build a brand around his personal name anymore.

Examples

• Rand Fishkin. Known for Whiteboard Friday SEO content made at Moz.

• Gary Vaynerchuk. Known for Wine Library TV. Now has Empathy Wines.

• Peter Askew. Known for building Onions.com, became popular after the viral article I Sell Onions on the Internet.

TLDR: Ryan focused on making things that people liked, rather than building an audience. That grew his following, skills and assets. He then used those to build more things that people liked. He rich now.

Focus on the work and become known for something, not about something.

At the risk of adding waaay too many links, here are a few other bits I’d like to share this week:

1️⃣ How to run smarter LinkedIn ads for SaaS and services

I met , a sharp LinkedIn advertiser doing work for B2B clients - and had to fire up my podcast again to record a conversation. We got into targeting strategy, what works in LinkedIn Ads, and how to make LinkedIn actually generate pipeline without melting a credit card.

🎙️ LinkedIn Ad Strategy for SaaS & Services, with Mony Chhim (40 minutes)

—

2️⃣ A better way for video creators to showcase their work

My friend just launched ProudWork which gives videographers and editors a bloat-free video player that streams directly from your cloud - no uploads, no distractions.

🛠️ ProudWork.io

—

3️⃣ How I dropped Facebook lead costs from £8 to £2

I made a video breaking down a local lead gen account I turned around - cutting CPL from £8 to £2 and keeping it there. This isn’t a fluke week, either, thankfully. Might help if you're working on Facebook Ads for local or service-based offers.

📺 How I Dropped Facebook Lead Costs by 75% (10 minutes)

P.S. Want to build something together?

Here are a few ways we could collaborate:

• Validate your product idea: Use HypeDeck.io to get early feedback before building anything. Great if you're still testing the waters.

• Work with me on Facebook Ads: If you want help running ads or want to bounce around strategy ideas, reach out here.

• DIY your own ads: I made this for my friends who asked for a speedrun on DIY lead ads. It’s how I do what I do, condensed into a digital playbook. You can get it too on Gumroad.

Key links:

• Mony on LinkedIn

• The Marketing B2B Podcast

If you’ve done one, hit reply and send me the link - I’d love to read it.

/ Ben

It’s not the size of your audience but how focused it is

Nathan Barry explains how to get more results from a smaller audience. The trick? Focus on a niche that deeply needs what you're offering, rather than chasing bigger numbers. (This also mirrors something Alex Hormozi said recently about accounts with low followers but high conversions.)

✍️ The Audience Shortcut, by Nathan Barry (6 minutes)

How to prep for AI platforms closing the door (on you)

Brian Balfour breaks down how platforms go from amazing to unusable (“enshitification”) - and why AI tools will go the same way. This post offers clear strategy on how to build defensible distribution now, before things change.

✍️ The Next Great Distribution Shift, by Brian Balfour (8 minutes)

High agency: the mindset for actually getting things done

Not usually a fan of inspirational pieces but this was cool. If you’ve ever wondered why some people seem to move faster, this article gives a useful model. It shows how “high agency” people think - and gives practical ideas for building more of that mindset in yourself.

✍️ High Agency, by HighAgency.com (5 minutes)

How to choose and grow a SaaS product idea that will work

A tight 20-minute staple in the MicroSaaS community from a serial founder on how to pick the right SaaS idea - and what not to waste time building. Great for anyone who wants more structure or filter in their product thinking.

📺 How to Build SaaS Products That Work, by Tyler Tringas (20 minutes)

The media formats people actually engage with (by generation)

This graphic breaks down preferred content formats by age group - from newsletters to TikToks. Good reminder that your audience likely consumes in more than one way (you probably do too).

📊 How Different Age Groups Engage with Content, via NP Digital

How to get more GPT mentions by name (yes, really)

A short thread that explains how to make your name or brand more likely to be referenced inside ChatGPT’s outputs. If you’re thinking about AI-era awareness strategies, this is a cool (and actionable) starting point.

✍️ How to Get Mentioned in GPT, by Ian Nuttall (2 minutes)

Resource: grow a YouTube channel like it’s your job

My mate Sean grew his channel to 585k+ subs, and now he’s teaching others how to do it. If YouTube is your game (or might be), check these linx:

🎥 Sean’s Channel
🛠️ Work with Sean

P.S. Want to build something together?

Here are a few ways we could collaborate:

• Validate your product idea: Use HypeDeck.io to get early feedback before building anything. Great if you're still testing the waters.

• Work with me on Facebook Ads: If you want help running ads or want to bounce around strategy ideas, reach out here.

• DIY your own ads: If you're more hands-on, I made this short course to help friends get started with Meta ads. Take a look.

Well, this has definitely improved my quality of life.

But what’s more, I found I had more time and headspace to focus on the bigger picture. So since I’ve stepped away from building an agency, I’ve had more space to work on longer-term vision:

Creating a time leveraged lifestyle business.

Last year I launched Lulara (which I’d still like to test another 1–2 times before letting go), but I didn’t really know where I wanted to take things.

Then in March, I learned to code and built Growl, which opened up the SaaS path, something I hadn’t seriously considered before.

While it started as a learning project, I put it out to the media buyer community to see if it could be sold following the ideas of The Minimalist Entrepreneur, by Sahil Lavingia the founder of Gumroad, that suggested trying to build for your own community.

That’s when I ran into the same issue most indiehackers face:

Making a good product isn’t enough. The real test is making things people actually want. 😏

To solve this, the usual advice is to make a landing page and collect payment to validate your idea. But Growl wasn’t finished yet - so no social proof, no reviews. Just features. I was honestly a bit stumped on what to put on there, even being a marketer myself.

That led me to a new idea.

What if this could be as easy as making a Linktree but created something like an app store page - where users could give feedback on the idea itself before being created?

So I made HypeDeck, which helps makers easily create social feedback pages for product ideas to easily validate ideas (or features) before building them.

I got started with gusto building the MVP in Cursor, hosting it on Cloudflare Pages for free, and then shared it with the indie maker community on Reddit, Discord, and Twitter.

But after ~400 visitors and zero sales 🍩, I nearly gave up.

By chance I spoke to a friend about it and they LOVED the idea but wanted a few changes. So I added them and they ended up buying. Woop, woop! Then to avoid getting fully over my head, I partnered with my mate Zak to build, letting me focus on promotions.

Now I just need to promote it more from here and get a better beta version going.

Both of us are keen to see how far something this simple can go.

YouTube Stagnation

I’ve tried not to neglect my main business. So I revamped the business webiste, furthered my network, and worked toward the goal of posting 100 videos on YouTube.

However, earlier this year I saw a solid video on monetising your agency with DIY, DWY, and DFY products, plus up/cross/down-sells, thought it was brilliant and started building that out. And given that a few folks had asked for a course, I started making one.

But during the course creation process, I didn’t have the mental space to also keep up YouTube.

I just managed to squeeze out the course without going insane, and while the course sales were OK, this totally killed my YouTube momentum.

Thankfully, my mate Sean started a YT coaching business and has been a good motivator to put more chips in. So I came back, posted a few Shorts, and am back in action.

Considering this year’s goal ( hitting 100 videos ) it looks pretty unlikely for the year.

So I’m considering changing tactics: Focus on Shorts + a few quality, trust-building long-form videos.

That seems sustainable and more effective for now. Let’s see how it goes.

Newsletter Update

Thankfully I’ve been pretty consistent here. It’s easy to maintain.

The process is to I collect ideas, links, and thoughts I find interesting during the week, then stitch them into a weekly roundup. It takes maybe 30 minutes to and hour to write and send.

But ironically, it hasn’t generated any direct leads so far. Ha.

Ultimately I’m okay with that. I see it more as a networking and staying-in-touch tool. And since it plugs my stuff (and my friend’s work) occasionally, it’s still a worthwhile part of my marketing, at least for the rest of the year.

I might change up the format soon, mixing curation with a bit more writing, partly to scratch the writing itch.

Kinda, maybe getting disrupted by AI?

While I still enjoy doing Facebook Ads, I’m starting to think our time is limited. Not because AI will do everything, but more like how SquareSpace reshaped web design by taking what used to be $5,000 - $20,000 into a $50 monthly subscription. Perhaps eventually, only the top operators will survive.

So I’ve been investing in:

• Video skills

• Storytelling

• AI content generation

I’m also diversifying.

As you’ve probably noticed from this newsletter, I have a hunch that people who can use their work and life naturally as personal brand content will do well. So that’s the next skill I’m working on, and luckily I don’t need to launch anything new to do that. I’m just documenting what I’m doing now, like HypeDeck and building Hungry Bear, practicing creating content that reliably generates leads and sales without feeling like a total tosser.

I’ve made some progress, but it’s nowhere near where my ad skills are yet. Still, I’m curious to see how far I can take it.

Wish me luck for the rest of 2025. Wishing you luck, too.

P.S. Want to build something together?

Here are a few ways we could collaborate:

• Validate your product idea: Use HypeDeck.io to get early feedback before building anything. Great if you're still testing the waters.

• Work with me on Facebook Ads: If you want help running ads or want to bounce around strategy ideas, reach out here.

• DIY your own ads: If you're more hands-on, I made this short course to help friends get started with Meta ads. Take a look.

And smart people like the folks at Y Combinator tend to agree. They found that only 4% of the top 100 startups that went through it at some point were solo founders.

But what if you kinda hate partnerships?

I’ve partnered on over 7 different projects with friends (whom I quite like), while things do get a little better each time the same or similar problems crop up:

• Personal volatility. Some people are emotionally stable. Others aren’t. It’s hard to know until things get busy or execution is difficult.

• Unclear responsibility. I’ve mostly fixed this by having a clear partnership agreements in place, but sometimes it pushes into a grey zone where both parties expect the other to be caring about some thing.

• Lack of self-awareness & skill. Similar to volatility, we all lack self awareness to some degree so it’s hard to know whether Person A should really be doing some task or that Person B should doing it. What do you do when they love it (and are crap at it)?

• Lack of leadership. In some situations, you can agree Person A is the leader, but when it comes down to certain situations, are you going to push against that agreement in order to take the project in another direction that you think is better?

• Loot sharing. It seems counterintuitive to even partner on projects that are small in nature, since sharing the loot on something so small makes both people struggle, rather than a single person comfortable.

🌩️ Before you smirk ironically at the idea of me blaming everyone else but myself - hold on. I’m with you — I could be the issue. Haha.

Even so, partnerships seem like a weird way to do business. Especially for a small or indie project striving towards a sustaining lifestyle business, with plausibly lower rewards than going for a big time traditional startup.

It’s like hiring an employee and deciding to give them half-ish of your project that makes a freedom-level $2k per month into a ramen budget of $1k (practically less, after costs). Add on a personal relationship to list of things you could lose from this arrangement if you know the person…

But thinking aloud, what if there was a way to make it work?

What if small time founders just focused on partnerships when the effort was a clear 90 / 10 split in terms of work and responsibility. So instead of full time equivalent partners, contributors are asymmetric allies.

What I mean by this is partnership engagements where:

• The help is useful but non-critical. The core operations can still run if the partner sucks or vanishes.

• It’s modular. Their contribution is scoped and sliced and you can add multiple or replacement engagements of the same thing if necessary.

• Cleanly exit-able. By the way the infrastructure is used e.g. “Just promote us on your Youtube.” “Use our blog, but put your author name.”

For example:

• An influencer partners with an eCommerce operator. But the operator already has a running business servicing multiple people so the relationship has no stress if it works or doesn’t. Likewise, the product being promoted by the influencer is just a small part of their overall portfolio so it makes no difference.

• A marketer who partners only with a developer, where the coding required from the developer acts as a CTO to the marketer’s vibe-coded mess. The show could continue if the CTO partner leaves.

👉 You can see it goes beyond simply affiliate sales. It could be applied to a whole spectrum of skillsets.

While we may not get the key benefits of sharing a vision with a true founder, I’d argue we may be able to get the key benefits of partnerships without the drawbacks of being sewn in.

You’d still get:

• Emotional support, even if lesser than a regular co-founder. I know I’m emotionally invested in projects I contribute in, and even my client’s businesses! Why wouldn’t someone else be who benefits from the success of the project?

• Shared vision. I don’t know about you but smart people can’t help but give feedback when they see something going off the rails. Not a certain thing for sure, but there’s also a chance the feedback may be even better since they’re less emotionally wound up in the project.

Stuff I’m Still Figuring Out

• Money motivation without equity spaghetti. Revenue shares make sense for revenue-generating partnerships, like those with marketers. But what about developers or service/product delivery folks? Fractured and small equity stakes are probably not worth getting into either.

• What if you outgrow the model? I thinking that if the idea blows up, you’ll revisit roles anyway. A good problem to have.

• Legal formalities. I’m guessing you can write a special kind of contributor agreement: like two pages, plain English, auto‑vesting and winding down.

Anyway, that’s the idea I’m playing with. Basically, if you’re about to handshake a 50 / 50 split on a micro‑product, maybe consider the 90 / 10 view first. Worst case, you keep more pie and do more work. Best case, you dodge the partnership drama entirely.

On March 21st I posted an update sharing that I’ve started trying to learn to code, again, for the 3rd time. And that I hoped I could push through and build something as kind of learning capstone.

Thankfully, I managed to push through this time and created my first working app. Growl Copywriter (link). Even if it’s a crappy product, I almost can't believe it. I never considered myself to be the coding type as I prefer not to get into the deep details. You can tell by the amount of typos I make everywhere.

Anyway, if you’re a marketer considering learning to code then hopefully this post gives you some ideas or inspiration.

From Learning → Building

In my original post, I shared I’d try stick it through one learning course. Here’s what actually happened:

• 70% completed the Foundations course from Founder Hacker by Ryan Kulp. I almost finished the course, bar 2 major things (ironically probably the most important: APIs and object oriented programming). I didn’t finish them because I have a bit of perfectionism getting in the way because I want to really follow the tutorials and type out the code for myself, rather than just speed run the videos to make sure I really understood what was being taught.

• I watched a bunch of YouTube videos about MVC (model, viewer, controller) architecture. Understanding this took a bit of time. Somehow it seemed not that logical to me, like why are Routes not the same as Controllers, in function? And why even do it that way in the first place versus having a bunch of Routes and single files that combine Controllers and Models. I had to chat with Chat GPT a lot about this too to better understand.

• Got Ryan’s 24-hr MVP course too, then basically just followed along his videos but did it for my own project idea rather than his explanatory examples. This was golden because it walked me through the whole process of planning, building and deploying an app for real. For example, Ryan explained wire-framing, building databases, creating the front end, backend and deploying on Heroku. I literally just followed the same process.

• Used Cursor to help me through it. I can’t claim I wrote all the code, though. Writing that as a total newbie is super hard. Cursor helped talk me through it and write a lot of it, honestly. But just by having finished the app, I have a much better understanding of how things come together and how they work and I can assume if I do this 100x, I’ll have a really good feel for it.

• Started some basic promotions on Indiehackers, shared to my community of fellow media buyers (link) and annoying friends with it. I feel it has a chance at commercial viability, so let’s see. Build, measure, learn, right?

Looking back, this was a surprisingly effective way to go about it in my opinion. I didn’t get stuck in tutorial hell or anything and was able to produce something by the end of it.

Idea Selection: Why Growl?

Frankly, most of my project choices in the past have come out of my own curosities rather, except for my D2C brand where I did heavy market research first.

In this case, I was heavily inspired by Pieter Levels book MAKE (link) which suggested focusing on solving your own problems.

This was interesting to me because it seemed to be backed up by two different authors:

• Sahil Lavingia who recommends building products for your own community in his book the Minimalist Entrepreneur.

• Paul Graham, YCombinator guy - see the infographic below:

Reflecting on this, it seems like a good move compared to more abstract market-research driven product selection because I’d have:

1. A better understanding of how the product should work, and why I want it.

2. Natural connections to people who might use it or give feedback.

3. A bonus of community-kudos for making it, even if it fails.

Indiehacker Oliver Eidel wrote a cool exploration of this idea, too, in his article about becoming a full-stack person (link).

The drawback - I now get the apprehension Indiehackers have about sharing products though. When it’s your baby, you don’t wanna drop it in a chat/group etc and have people dunk on it and tear it to shreds. Funnily enough I don’t feel that way about videos or blog articles but it’s more severe with a hand-made product.

Learnings About Coding

Including this for fun reading for actual programmers, too:

• Libraries and shortcuts are cool? I had no idea using these kinds of shortcuts or outsourcing solutions for things like authentication was so normal, practical and even cool in dev circles. I thought programmers would just write everything out from memory given they must have done it a million times, but that was obviously wrong. lol.

• There’s a surprising amount of terminal checking. I realised if you save testing for when you think you’ve finished a feature, the whole thing can be stuffed up and you don’t know where. So I eventually got into a groove of testing everything I could in terminal as I went along so you don’t end up at the end with a broken implementation and not know why.

• Plan file & folder architecture ahead of time. There were a few occasions where halfway writing code for something, I realised I should have structured the code a different way. So I learned that you better think through how you’re going to structure the whole program (files and folders) ahead of time to your best guesstimation to avoid having to redo stuff later. It also seems to help with making the program more scalable. E.g. making features modular in style, so you can simply plug in a new AI or more inputs, etc.

• I learned what it meant to be opinionated with Ruby on Rails. Small thing, but I got a little bothered (probably unecessarily) that Ruby on Rails came with so much stuff, that I didn’t know what it did, in the beginning. And that it wasn’t really made to be flexible in how you made things.

• Be very careful to avoid security flaws as you go, like SQL Injections, XSS attacks, authentication, etc. My mate Alex kindly took me through how to some issues here as it’s a common issue for newbies and people who use AI to code.

• You can’t really code without the internet? I commented to my mate David that the internet went out once and I was pretty useless. He said that’s normal. Again, I assumed most programmers just write things in a kind of finger-tip native fashion, haha.

Overall, I’m pretty thankful to have friends like Alex for helping check through and Jesse (Indiehacker in Bangkok/Philippines) for guiding me through the learning journey.

Ultimately, I’m not very confident at codings till but I feel like it may be similar to marketing where you’re always trying to improve and add things into your techniques and get better over time.

What’s next?

I have some hardware projects I’d like to do that would use MicroPy, so I’m looking to use my next project idea as a chance to try Python and A/B languages. Let’s see if I like it overall for building versus Ruby.

I’m also going to continue promoting Growl for awhile, see what I learn and if it becomes it’s own thing.

Otherwise, I think that’s it for this series. Thank you for reading.

In this post I’m reflecting on the lessons learned from starting, managing and growing this group until the hand over. If you’re considering running a local technical group this may be interesting to you.

🚀 Starting Up

Some thoughts about what assumptions we had that worked out (and those that didn’t). And interesting benefits we found out about later:

• It is indeed a super fast way to build a network. Starting out, we kinda just wanted to make new connections and learn a few cool SEO strategies. And for this purpose, the group was fantastic. I’ve met some of my closest friends through the group.

• Facebook beats telegram, probably. We decided to build the group on Facebook instead of a chat channel like WhatsApp or Telegram. And I think that was the right choice. Looking at other groups run on chat apps, the conversation seems too fast to properly notify people of events, which you probably need to post somewhere like Facebook anyway.

• Two hosts are better than one. Running the groups is way easier with two hosts instead of one. You get sick, travel or sometimes just want to do something else. Amund and I took it in turns to run the events themselves, with him doing most of the hosting when in the country and this worked well.

• It’s a great introduction mechanism. Bangkok is one of the most visited cities in the world, especially for digital nomads. Having the group was a great way to introduce new friends to more new friends in a very quick, social way. If someone visited me, I could just bring them along to the group and they can meet and make their own friends there in a single night.

🌱 Growing the Community

TLDR, organic growth was the main method but there are some nuances:

• Start with your own friends. As the group started, each of us mainly invited existing friends and connections in Bangkok. That was a fast way to get the original group going.

• Promote in other friendly groups. Another technique Amund used was to promote the group and its events in other friendly groups (with permission), like the Bangkok digital nomad groups and the Chiang Mai SEO Community itself.

• Meetup.com was a waste of time. Not because it didn’t get attendees, it did. But for this particular group you needed some hands-on experience with SEO. Meetup.com attracted a lot of generally curious people, who found it hard to integrate with the SEO nerds that typically turn up. Had the group been a bit broader in context, like generally being a digital nomad, it would have been a good growth channel.

• Organic growth is the 80/20 driver. Eventually, the group took on a natural, slow but steady organic growth trajectory of it’s own. We didn’t have to try grow it as people brought friends and their friends brought friends.

• You need to have a good consistent time + location. This wasn’t one that I originally thought of, but the location and consistency of events really made a difference. It’s intuitive really. Things started to grow faster once we picked a consistent time, the 2nd Wednesday of every month. And also when we changed venue locations, too. Amund spent a huge amount of time trying to find a good location and that was really worth it.

📋 Managing the Group

Managing the group was the least fun aspect of it, so probably more hard lessons learned here.

• Don’t forget to add the actual rules to the group. We forgot to add rules to the Facebook group, and eventually some people started spamming it or behaving poorly. It was a bit awkward to remove/moderate those folks without having rules to point to so don’t forget to add those early on.

• Establish a rhythm for the event, e.g. eat first, chat after. In the beginning, it was a little unclear for us whether we were there to eat, drink or just chat (or all 3). Towards the end there it because clearer that we’d all meet for dinner then chat/drink after. Setting this expectation seemed to improve the smoothness of the events themselves.

• Don't feel stressed about people having a good time. In the beginning, I spent a lot of effort trying to make people feel good and welcome at the event. But somehow this made the events feel like much more hard work than letting it flow, whatever happens. Loosening up my own expectations here made it easier to run the event. Ultimately, if people can socialize - they should. If they can’t naturally, it’s not necessarily your job to help them.

• Deal with the trouble makers through key members. This was an interesting one. With every group, there are dramas and accidents. Because we grew predominately organically, we had the benefit of asking members who knew the person who created the drama to help manage the situation. That worked great.

🚪 Handing Things Over

I had one false start trying to find new leadership and management for the group, but Amund eventually found another organizer to take my place.

• Make an email list. We didn’t, and should have. Having an email list would have made the group more valuable to sponsors and notifying people of events easier, but sadly we never made one. Not sure why exactly. Probably just lazy.

• Consider adding potential hosts in early. The replacement host for me was an active member of the group already. It was a natural fit to hand things over. So if I went back in time and needed to find new leadership, I would have just invited one or two people was assistant hosts for a few events then pick the best one. Benevolent dictatorship…?

• Perhaps university-style club management practices are a good idea. It’s too late for the community now, but it would have been interesting to see how the group would have operated and grown had it been run like a university club. Yearly presidents, treasurers, events calendar, website, email list, etc. These are probably why university groups live for so long in the first place.

Not because I want to become a developer or change careers. I just think it’d be cool to understand enough to build my own little projects - things I’d actually use. Maybe even see if anyone else would pay for one someday.

Attempts So Far

This is technically my third attempt.

First time, I watched about 4–6 videos from Network Chuck on Python. Pretty easy intro but it didn’t really stick. ended up not continuing from there.

Second attempt, I downloaded a bunch of Ruby books on Kindle thought I’d absorb it that way because I’m more of a reader. But that didn’t really work since it’s an even more dry way to work through data types, functions, etc completely without context.

For context, this isn’t completely from scratch. Back in the day, I used to mess around with custom servers in Graal Online. Learned just enough scripting to spawn tanks or special weapons. The syntax is reportedly vaguely similar to Java or C, and somewhere along the way, “if” and “then” logic kind of stuck in my brain. Not that it helps much now, but it’s there.

Motivations

Honestly, part of this post is so I can look back one day and see if I actually broke through or if this ends up being another failed attempt.

Here’s what’s driving me right now:

1. I want to try build some things. Probably just for myself, but maybe someone buys one. Who knows.

2. Needed for other fun projects. Some of my future project ideas lean toward hardware/IoT stuff, so I figure this is going to be necessary.

3. General thinking. Perhaps learning to code adds another dimension to how I think about solving problems.

The Plan

My goal is to learn by doing. Push through and ship something using the skill.

I will try follow a course, but knowing myself, I’ll probably bounce around, taking in different little things as I go. I will hopefully journal the progress here, too. I doubt anyone reads it, but if someone does, that sense of being “seen” helps, you know?

For the language, originally, I picked Python because I liked the overlap with ethical hacking, AI, and potential IoT projects. JavaScript and Go got recommended a lot, but neither really took my fancy. [Later: I changed to Ruby].

Progress

• I watched Network Chuck’s Python Series. Maybe 6 videos. Gave me enough momentum to start but but eventually I ran out of interest/steam. Didn’t finish it.

• I kinda let this project go and forgot about it until I saw a video pop up on YouTube called Animation vs Coding (above). It showed a stick figure visually demonstrating python code. I thought to myself, “What would I feel if I just copied everything line by line?” So I did. Took me like 2 hours. Really gave me a contextual boost, I think.

• Ended up making my app work with Replit but then I talked to a friend at the office (thanks Jesse!), who mentioned because I used AI had I had rammed everything into a single .py file. So I couldn’t really work out what to fix etc. Made me realize I probably should learn more foundations or at least code structuring before vibe coding my way into trouble.

• Got Ryan Kulp’s Foundations that teaches Ruby for free - March 4th, which in the absence of a clear plan (of learning how to structure code) I started sort of chipping away at.

• Now I’m fairly motivated and find myself sneaking away in the mornings or trying to add a few minutes of study time for Ryan’s course in the evenings.

• I’ve also been trying to cement my learning so far planning hypothetical apps in Chat GPT and asking if I’m understanding how MVC and file/function and data processes work together. I’m getting there for simple ideas. Just can’t write up everything together yet.

A quick tally of learning time so far:

• Network Chuck’s python learning series (link), probably 6 videos (each 20min) + 40 min practice and follow along time = 6 hrs.

• Copying stick figure video mentioned before. 3 hrs.

• Reading book on Ruby, 0.5 hrs.

• 40 lessons out of 89 in Ryan’s foundations course. Probably 20 hours?

• Q&A with Chat GPT, 4 hrs?

Total approx 33.5 hrs.

Early Observations

• So far I’ve noticed learning really is messy. Learning doesn’t seem to happen linearly where you follow one resource from level 0-1. I’m sort of bouncing around material a bit which keeps me engaged and somewhat motivated which leads me generally back to Ryan’s course.

It’s funny how common this is. I realised it’s the same with other things like learning languages. Sometimes you just gotta throw everything at a general direction and pile up trash to build a mountain.

• It was interesting to see how pseudo code was a real thing. When approaching problems, I instinctively wrote out what I wanted each piece of code to do in English then tried to translate that into working Ruby. After awhile, I got more of sense of what was actually possible and how things worked, so I got closer to getting the approach right

• It’s pretty interesting how tactile coding is. I didn’t really expect it, but physically typing out everything being demonstrated makes a huge difference to seeing it work. really cool and I wonder if I can improve my learning potency elsewhere by doing the same.

• I’m surprised at how coding involves so much Googling! Lol. I somehow had this idea of coders who mostly wrote things from memory, basically crafting solutions and laying pipes with math (lol?) straight from their head. but it seems like there’s a big element, or at least for me, of Googling different things for the language on what methods to use or fixing problems of using those methods.

Anywho, thanks for reading this. Let’s see if I manage to push through enough to eventually make something.

Motivations

• Diversify my income. The issue with my current work is that when I get unwell, there's no one to handle client work for me. I've tried training people, but it's been super difficult. I figured the problem here was that ads management is a difficult product to deliver, whereas eCommerce is mostly operations (from my experience).

• Physical products vs digital services. I’ve also wanted to move away from digital services as the only thing I do. I’ve worked with hardware in the past and loved it. I often wish I could get back to doing more with physical goods.

• Learn by doing. While I had some level of success doing eCommerce back in 2013, it’s been a long time. I so wanted to do a "full rotation" of practice. Just block-headedly go through all the motions to complete a project, even if I know that the project is going to be completely crap - to learn what needs to be learned rather than getting stuck in over thinking zone.

I got a huge boost in motivation here from my friend MotoMatic who has great success in developing and marketing his own designs. He encouraged me to not be the midwit over complicating it (above) and just push through and learn.

Choosing a Product

The first two months after registering my company were spent considering importing and distributing a friend's Korean cosmetics brand. My rationale was solid—I'd seen unauthorized resellers already operating in Thailand, indicating demand, plus I had a direct connection with the founder.

One of my team, Rei, did most of this and a fantastic job. However, after researching all the required processes, we discovered it would cost around $6,000 USD due to Thai FDA approval requirements and specialized warehousing needs. That was a non-starter.

↔️ Weighing my options, I remembered the primary purpose of this venture was learning. So I pivoted to creating my own brand instead.

I'll admit I got a bit idealistic here. I wanted to sell something that provided net positive value to the world - not just trinkets and junk. With a friend's help, I profiled Redditors' favorite purchases and decided on sleep products, starting with a sleep mask to sell locally and abroad.

The product didn't need to be a home run at this stage; it just needed to be a strong exercise in the entire process to gain that experience and understanding. So I committed to completing the process with this category and product.

The Design & Manufacturing Process in Thailand

I used the "double diamond" design process to guide my research, development, and manufacturing approach:

1. Research & Analysis: I analyzed feedback about sleep masks on Amazon (reviews of top sellers), Shopee, and Reddit to understand which designs people liked, disliked, and why. This gave me probably 1,000+ data points to work with.

2. Market Research: I analyzed price points locally and internationally using Shopee and Amazon.

3. User Definition: I defined my target user as women aged 30-40, motivated by health and beauty concerns.

4. Problem Identification: I established six specific "problems" that my product would aim to solve, based on the review data points. I’ve got it stored in a Chat GPT log somewhere…

5. Competitor Analysis: We ordered and reviewed 15 sleep masks from existing OEM suppliers to understand the market. We took turns testing each at night and checking which ones we liked and didn’t like, what the problems were, etc.

lulara.sleep

A post shared by @lulara.sleep

6. Prototyping: We established a first prototype vision - not a sketch, but more of a "Frankenstein" of existing products with improvements. Basically, the best part of this with the best part of that.

6. Manufacturing Partnership: Rei contacted around 10 local manufacturers specializing in textile and foam products. One local manufacturer (a personal contact) agreed to create our design pattern and prototype.

The pattern maker from the factory basically took a list of our requirements and preferences and made the prototypes form there. They were very gracious and patient with us during the process. And allowed me to make several designs with them.

Basically we’d give them a numbered list in a Google doc of what we were looking for plus a visual example of the thing, if possible.

Initially, the results were pretty discouraging. I felt like I was never going to make it and I was just wasting my time while making a fool of myself. But I allowed us to keep trying a few more prototypes and they eventually started looking much better.

Finally, we got a prototype that looked good enough to present to real potential buyers. Initial tests with people in our office was promising - our design was preferred (in style) over all the other masks we had purchased.

During this process, Rei and I kept cost a feasibility analysis going with conservative figures, initially assuming Chinese production. The target numbers were based on a net 30% net margin goal, including all selling and marketing expenses.

But this led us to hit our first major obstacle: the cost estimate for minimum order quantity (MOQ) came out at $8 USD per piece, significantly higher than our maximum target cost of $7.50 (for a $35 USD sale price). So it wasn’t going to leave much, or any substantial margin left at that point.

Also, this prototype had a major flaw. It was too hot when worn for extended periods. This seemed to be related to the fabric we used as a "blackout layer."

So we requested the next version with a replacement for this component which seemed to fix the problem. Phew.

Eventually, we got production costs on target too by cutting down on other luxurious, ensuring each component was priced properly to maintain a 30% net margin.

We received three more samples from the Thai factory to confirm some minor changes and color options, ready for photography. These were preferred by our office testers 4 to 1 against existing designs from China. Finally something we could run with.

Or so I thought.

The Quality Roadblock

Looking more closely at our samples, I noticed inconsistent quality between them. The velcro broke on one, the nose cut was inconsistent, and the finishing at the back wasn't smooth. It was difficult to commit to a full order knowing that perhaps 30-50% could turn out subpar or break during normal use.

(And a good thing we made this call too. A friend later mentioned that the nose degree on masks in China is allowed to be only something very small, like 1%. Whereas our supply chain here locally seemed to be not suited to meeting the same standard).

I felt pretty discouraged until I remembered my friend’s advice to just go with an existing Chinese design and supplier, then iterate on a custom mask later in the background. The project was taking too long, so I bit the bullet, changed track to a Chinese supplier, and essentially restarted the whole setup process.

From basically 0 again.

We sourced three existing designs from Chinese manufacturers, though none were particularly impressive at first. This was a significant setback, but we managed to switch quickly to a reliable supplier who could meet our quality requirements.

The delivery of our minimum order quantity (100 pieces) came to Thailand and was surprisingly painless. We also secured a local Thai supplier for the mask bags and boxes, which turned out much nicer than expected.

We also had a local supplier for the mask bags which was easy to deal with and made in Thailand. The boxes also turned out much nicer than expected, too!

Unfortunately, not everything survive the transition from idea to reality. the fancy packing paper and stickers we had designed proved impractical, resulting in a small waste of time and money.

Building the Brand Identity

Creating a distinctive brand identity was a crucial part of the process:

1. Naming: We decided on the brand name "Lullara" (pronounced "Lu-lah-rah"), derived from "lullaby." Later, we simplified it to "Lulara" (Lu-la-ra) for easier pronunciation and secured the domain lularasleep.com.

2. Brand Character: We developed a mascot character named Lara, an otter, to improve brand distinctiveness and engagement.

3. Visual Identity: We created mood boards to explore the look and feel of the product and branding style, working with a local Thai designer who developed logo concepts.

We settled on a style that I hoped conveyed a premium, cozy, and easily recognizable look:

4. Packaging Design: And we finalized the pattern for the packing paper to make the unboxing experience cool.

During this period, I also did the basic setup tasks:

• Set up Stripe payment processing PayPal.

• Established a bank account and discovered that business debit cards are uncommon in Thailand during a company's first year - in fact, some people still don't have them after a few years!

• Onboarded with PEAK accounting (a Thai version of Xero).

• Created an official LINE account for customer communications.

Product Photography

We A/B tested photographers in Thailand from Fastwork (a local Thai version of Upwork). The cost was extremely reasonable - approx 2,000 THB ($60 USD) each for around 10 photos.

We provided them with a detailed brief including a shot list (in Canva, I think?) with examples and background materials on the brand to establish the desired look and feel.

A key learning here: photographer style matters tremendously. Male photographers seemed to struggle with capturing the "feminine" style we were aiming for, so we received some questionable shots.

However, we also got some excellent photos that captured our brand aesthetic a bit better.

I got help and then spent half a day putting everything together, packaging about 65 out of our 100 units into the final product presentation.

Warehousing

In the beginning, I wanted to go with a 3PL warehouse so that I wouldn’t have to handle stock. Sure, it might flop (and it did) but if things did work, I have a full-time role managing Meta ads for people so I didn’t have time to run around posting stuff.

But ultimately, someone from the team volunteered to handle product storage and fulfillment (pick, pack, and send) in exchange for a prorated fee comparable to what we'd pay a third-party logistics provider elsewhere. So that was a win-win.

To pack the products once all the components were delivered, I got help so it ended up being about 4 of us trying to assembly everything. It took wayyy longer than expected, so we only managed to do about 65 in half a day.

But at least it was stored and ready to sell!

Launch and Promotion

We made our social media accounts across TikTok, Facebook, and Instagram, where we tried some "build in public" style posts along the way to warm things up. One post garnered 15 votes, which was encouraging.

lulara.sleep

A post shared by @lulara.sleep

We also looked up and reached out to 100 influencers in Thailand to build relationships for future product promotion.

Unfortunately, the manufacturing delay meant we had a short window to complete everything before the major sales period, and we missed it. This taught me an important lesson about timing in retail.

Another unexpected challenge: Lazada (the 2nd? biggest Thai eCommerce platform) has a cooldown period of 3-5 business days while they confirm your legitimacy before you can promote anything. This includes allowing tracking via Google, Facebook Ads, etc. This meant we couldn't gather reviews before launching, though we did go live in early December.

Despite these setbacks, we managed to get 15 influencers to promote the product (in exchange for free product), with approx 5-8 more on the way. Some influencers generated good engagement (likes - see above), though the impact on sales didn’t really come through.

We got 1 sale. And some expressions of interest.

But ultimately the feedback we got over and over again was the price (1,200 THB) was far too expensive.

Originally, I was going to do my own ads on this, though I’m not an eCommerce advertiser, but unfortunately in 2024 I basically burned out and needed to simplify what I’m working on for the time being. So for now the project is paused here.

Costs

Annoying for this article to not give more detail, I know, but I calculated the costs that went into the materials, designs and materials and it was about $1,500 USD.

May update this once I find the final spreadsheet. Doesn’t include labour.

Reflections and Lessons Learned

This has been a LONG process, but as an Aussie friend here mentioned, "not that long, you whinger." Here's the timeline:

• May 2024: Company registered.

• Months 1-2: Investigated importing an existing brand (later abandoned).

• August: Decided to "go through the motions" of creating a DTC brand to see what I didn't know, update my experience in eCommerce, and test some ideas.

• Late October: Switched to Chinese manufacturing, restarting the sourcing process.

• December: Launched in Thailand via Lazada.

What I Got Right

Some of my initial hypotheses proved correct:

1. Operating from Thailand IS cost effective AND can produce quality work. Thai people are an under-rated outsourcing talent group.

2. eCommerce is indeed easier to run semi-absentee compared to service businesses.

3. Rei has effectively run the whole show with my direction and funding - I've been focused on getting my act together on the ads business. She’s brilliant btw and available for hire, so email me for the connection.

4. eCommerce fulfillment is much simpler than client services (so far).

5. Physical product problems are genuinely more fun to deal with than client service issues.

What I Learned

But there were also some hard lessons:

1. OEM vs ODM. It's MUCH easier and faster to pick and customize an off-the-shelf OEM product than to design your own (ODM)

2. Product selection. As many experienced friends mentioned, product selection is exceedingly difficult.

3. Test, test, test. Testing quickly as a means of product discovery is a skill I need to develop in this business.

4. Inventory risk. This is a serious downside - you don't want capital tied up in unsellable or slow-moving stock as one of my friends warned me.

Looking back, it's clear that unless I already had a winning idea somehow, creating a custom design with fancy packaging and pieces wasn't worth the trouble. While I'm not comfortable selling products I don't 110% believe in myself just to find a winner, though I'm working on that mindset. Maybe I can or can't get over that?

Similarly, developing a brand before discovering a viable product was backward. Having a "nice" brand actually made me less comfortable about potentially tarnishing it with aggressive product discovery experiments. What a error.

What’s next?

I’m happy that I did this project and I have learned an absolute tonne. It’s kinda a wimpy way out, but I did sorta achieve my objective of learning by doing? Just… didn’t really sell anything with an organic-only launch. Haha. 😂

So I’m weighing my options now. I could…

1. Invest the time in another thing. Sell the brand if I can.

2. Take the existing product try a different market? E.g. US, Australia, etc. with better spending power.

3. Try a different product in Thailand with the same brand with a much, much lower price but more promotion via ads, organic, etc.

So if you’ve read this far, what would you do?

Since December 2020 I’ve been learning the Thai language. While I’m no where near proficient yet, I’m at the point where I can have some conversations beyond ordering food at restaurants with people who don’t speak English.

If we assume Thai is similarly difficult to English, then this Preply diagram based on the EU language levels suggests perhaps I’m at A2, though I manage to stretch the limited vocab I have to around B1:

Anyway, I’m writing this before I get too deep in my journey and let another 5 years pass without reflection. I want to do a quick recap of my journey so far in case anyone wanted some ideas on how to go about learning Thai themselves.

Here’s my learning journey recap:

1. Motivations

Everyone has their reasons. These are mine:

• Survival. Accidents and fights happen (verbal). I’d like to know enough Thai to be able to handle myself when things go wrong, without someone needing someone to speak on my behalf. Also helps that you can’t get tricked so easily if you know what people are talking about.

• Making friends. Learning Thai would help with making local friends.

• Competition. An odd one but motivating. In 2017 when I got to Korea, many of my close friends started learning with me around the same level. 4 of them, out of the group of 5 including me, reached TOPIK6 level in Korean, the highest. 2 of them even ended up on TV. Somehow, I didn’t want this to happen to me again in Thailand with the Thai language.

• The anti-goal of not becoming a frustrated farang at a bar somewhere. I’ve always had this in my mind to NOT become some kind of creepy uncle that scares kids at weddings, stuck in an unintegrated, low-achievement farang bubble. Withering away in a Pattaya go-go bar…

2. Timeline

Here’s a recap timeline of how my learning journey went.

2020

• Traveled to Thailand during lockdowns and wanted to start learning the language from knowing nothing about it.

• Bought a starter pack from Tutoroo (link). The learning was super inefficient and ineffective but let’s count it anyway. I paid 750 THB per 1.5 hours lesson, 2 lessons upfront. Renewed maybe 2-3 times? So I guess that’s 9 hours.

• I was in and out of Korea every few months this year.

2021-2022

• Basically gave up on Thai for the most of 2021.

• But in December 2021 I enrolled with Rapid Learn Thai (link), buying 2 rounds of 10k THB at 500 THB a lesson. 40 hours.

2023

• Felt a renewed focus on learning Thai.

• Did 200 days of daily study, before breakfast while exercising. Probably the minimum effort on Memrise which I think is 15 minutes. 200 x 15 minutes / 60 = 50 hours.

• At some point, I was reviewing vocab on the train on the way to work. I’m guessing I did this for about 3 months, each workday. 20 work days a month x 3 months x 10 minutes = 10 hours.

2024

• Moved to Chiang Mai for 3 months and made a routine to study about 30mins every morning with breakfast, each work day. I assume I missed about 1 day per week by being sloppy. 12 weeks x 4 days x 0.5 hours = 24 hours.

• Spent 6,745 THB for 20 lessons, 1 hr each. 20 hours.

  • I assume I spent at least 30 minutes study per hour with the tutor. Yes I’m a lazy student. 10 hours.

• ⭐ This is when my Thai really improved. I changed my technique to just speaking with my tutor for 1 hour about anything, using English only for words or expressions I didn’t know. Any words I didn’t know were written down for study later.

2025

• Bought a book, Complete Thai Beginner to Intermediate Course (link). Bought to see if it can fill in the gaps of my learning to date. But unsure if this will work yet.

• 4 dedicated study sessions in Jan/Feb (ugh, too few). Approx 30 minutes study time. 2 hours.

• Not yet in the habit of studying vocab on the way to work again, but probably done this 2x at 10 minutes each. 20 minutes.

Total stud time: approximately 165.3 hours.*

*Obviously doesn’t include passive time spent listing to Thai being spoken around me while being in Thailand.

3. Learnings

In the early stages, a structured introduction class would have been a more efficient way to learn the base layer of the language.

From the start I knew I couldn’t commit to a class based on my experiences learning Mandarin and Korean. My schedule often changes and unfortunately I tend to get sick often and easily. So studying only with tutors was my way to make learning flexible.

Sadly, this turned out to be one of the major things that made my learning time less efficient. Many tutors lacked a robust beginner learning process which would have been super helpful to get started. We’d cover a few key phrases but it didn’t feel like it went smoothly or properly built upon each other. It was more like going through a Lonely Planet guide of key phrases.

While I got through the total beginners stage eventually, it could have been faster. I still couldn’t have done it via a class, but maybe you can.

Learning reading early is a strong learning multiplier. But it’s just a multiplier, not a replacement for learning vocabulary and grammar.

In my early lessons with my tutor, I really felt like I wanted to learn more reading and writing as went.

I was worried I’d make the same mistake as I did with Mandarin, where I was guided to learn speaking and listening first - and would eventually come around to reading and writing later. That was a big mistake because while I could speak and understand (to approximately HSK 1.5 level), it was less useful because I couldn’t read basic signs in the Subway, restaurants or other places in daily life. Which sucked.

Having known that, I saw Rapid Learn Thai recommended in some Facebook group as a way to learn reading and writing first so I enrolled.

Gary Orman, the creator of the Rapid Method, makes several good arguments for learning reading first, too:

1. You can understand tones, so you speak more properly

2. You can read common words and naturally acquire more vocabulary

Looking back, I’ve found these 3 points (include mine above about reading in daily life) to be true in real life. Learning to read does make your pronunciation better because you’re reading the tones correctly. And reading helps you pick up words you see around the place, like หยุด! (Stop!).

One key problem I noticed though was that even though I was studying reading and speaking with the tones, I didn’t really know… anything? I couldn’t make sentences and I didn’t know common words. This actually got me into trouble once (ask me privately if you want to know how).

That’s when I realized that learning to read and pronounce Thai was just a speed multiplier for other learning. A nuance that I didn’t think of before I started.

Don’t forget energy management as part of your learning plan. Learning when I was tired probably reduced my efficiency 85%.

This was a big one. Because I could schedule tutors whenever I wanted, I often took lessons between things at work or after it. The problem was, I was usually completely toasted by the time I took the lesson.

It felt like nothing stuck and I couldn’t wait for the lessons to be over. I liked all my tutors, so it wasn’t their fault, I was just a crap, demotivated student.

It would have been better to schedule lessons only when I was feeling fresh at the beginning of the day. Sadly that hasn’t been practical for me, but it may be for you!

My number 1 study tip is to develop and refine a study habit.

Wrapping this up, my final lesson would be that the most important thing I did to improve my Thai was to develop refine a study habit.

First I thought of all the moments of the day where I could tack on some Thai study. Then I tested different methods to see what I could stick to. Over time, I’d try adapt my routine so that I did that every day, work or no work. This was how I got in two massive chunks of consistent study of 40 and 24 hours respectively.

Interesting resources:

• Journey: The Language Learner's Ultimate Guide to Maintaining Motivation by Erin N. O'Reilly (Link). I talked to Erin once for a strategic language learning consultation and she really, really knows her stuff.

• Language Lords on YouTube (Link). This dude is a machine at learning languages and has some really interesting, and motivating, techniques. Check him out.

So they head to Upwork, write a two-page job post (with the word "ELEPHANT" at the top), and start filtering applicants.

But this can often lead to deep, grinding frustration.

The freelancer doesn’t deliver what they expected. And I don’t mean the usual problems with dropkick freelancers, shoddy work, or bad comms.

I mean projects stall. Credit card gets melted. And you feeling like “Ugh, nothing works. What do I even have this person helping for?!”

As a freelancer who’s also hired a tonne of other freelancers, I’ve seen this happen countless times. And it’s happened to me too. But it can also be avoided by being ready for the freelancer, having the right one for the right engagement.

Here are some of the things that go wrong when these aren’t in place.

First, marketers aren’t product builders. So if your product isn’t finished yet, you’ll both just end up upset.

It’s hard to scale something unfinished.

• Running ads on a product with no product-market fit? Burn money.

• Driving SEO traffic to a business that can’t convert leads? Just frustrating.

• Booking outbound sales calls without scripts or follow-up? No one's closing deals.

Marketing freelancers deliver ROI if the foundation is in place. Otherwise, you’ll just spend money faster on a broken system.

Some problems are multi-faceted, meaning the person you’re engaging can’t solve everything with simply better promotion.

Wicked problems can be defined as those that have multiple dynamics of cause and effect, and are not simply solved linearly. While some marketers have great skill, solving wicked problems is more of a co-founder’s skillset than the remit of a freelance marketing guy/gal.

For example, a conversion rate problem on the home page is obviously a CRO job. But could it be a problem with your product? Or is it being put in front of the wrong market?

It’s worth considering that broader problems will require multiple sets of eyes to really crack, together with an open mind.

50% of the solution is cooperation, patience, and focus to see things through until momentum builds.

A freelancer being able to work independently should be expected. But it’s not realistic for them to work in a vacuum. Especially if the business can’t focus on either providing the necessary investment in the service, or the focus and patience to see programs and campaigns through.

You see entrepreneurs laugh about this all the time when giving each other advice:

“You sent 50 cold emails, and now you want to give up?”

“How much did you spend on ads? $150?!”.

Etc.

But it can happen to anybody.

Finally, if you want a generalist and to get things off your plate, get an employee.

Interestingly, one of the most common expectations I encounter is clients looking for generalists who can do it all. A very practical idea. But often, an impossible task for a freelancer who needs to specialise to earn a living against their peers in an open market.

Consider a regular marketing employee instead and task them with taking on multiple projects at multiple levels to see them through. While it’s not realistic they could write expertise-driven long form blog posts on aerospace technology for the company blog, they could coordinate the monthly blog process, scheduling time with you and a freelance copywriter, and editing the work.

A great in-house generalist is your quarterback. I don’t know anything about American football, but what I mean is that they replace your resource coordination, momentum-building, and review (or as much of it as possible).

Finding Problem-Person Fit

With that said, here are some thoughts on who and when you should hire freelance marketers for common problems at startups and SMEs:

• To offload a routine-ish specialist activity, like SEO, email marketing or CRO, pick a technical freelancer. Use track record as a reference for price. The better the track record, the higher the price will be. Make sure you have the processes that come before and after their part of the marketing funnel in place first. For example, having trained sales people ready for more leads.

• To build something defined, like an appointment booking process, website, email automation sequence, or brand identity, freelancers/consultants would be a good match on a project-based engagement. Example, a brand consultant, operations coach or web designer.

• To solve problems or improve performance, like conversion issues, brand positioning, look for specialist consultants. I think hourly consultations are the best value, versus a project engagement. Try to have relevant background information, clear goals and reference data ready to go and sent ahead of time before discussing the problem.

• To get things off your plate and to organise and run things without your input, consider a marketing employee. They are the human Zapier connector of your organsiation and a second brain to push things through. Find smart, organised and ambitious people. Bonus points if they have deepish skills in one technical discipline, so they can replace a freelancer in a key area.

I was Head of Marketing at a new education company and my forceful and fast moving boss insisted we test a lead generation technique by promoting a lead magnet using organic posts on the account.

Naturally, after finding out he got the idea from a course, I wanted to roll my eyes out of my head (FYI, we’re close friends now). I’d been managing corporate social media management for years, and from all the hundreds of posts across Instagram, Twitter and LinkedIn, I had decided that social was good not much. Just for updates and showing that the lights were on at a company. And courses? Pah, those mainly taught bogus techniques that were promoted as silver bullets to the masses.

But I followed orders. I briefed design, had the copy written and posted it. The caption and content asked people to reply if they wanted it. Then I moved onto the next thing in the day.

However, this time people responded. A lot of people.

But I didn’t feel relieved or anything. I felt.. annoyed.

Not because I was wrong in thinking it wouldn’t work. It always sucks being wrong.

I was annoyed because I realised I wasted years not even trying to be right.

For 9 years, I created copy, graphics, and scheduled posts - and I hadn’t properly learned anything about social media. Because I simply went through the motions, I didn’t know or care to find out:

• How people actively grow followers without more content, in sustainable ways.

• What types of posts were best for getting followers versus nurturing audiences.

• Image formats or copy structures for engagement or conversions.

• How to use call to actions in the caption copy or graphic to draw attention to a lead magnet or offer.

Most of these micro-tests would have taken just 20 min to research and test. Which could have built up slowly over time into a spell book of well-honed skills and know-how.

Pretty much every client or employer I worked with has wanted to do social media marketing, too. I have had countless opportunities to practice with live accounts.

So when the results were rolling in bringing new sales for the company it felt like a run-up kick to the pants.

This approach changed how quickly my skills developed. For instance, in my Facebook advertising the way I write ads in 2025 is way different to the way did in 2023. Why?

Because I’ve tested:

• Different ad frameworks

• Different ways of doing hooks

• Long vs short copy

• Carousels vs statics

• Researching positioning and appeals

• Implementing emotions into copy

• etc

Hundreds of iterations such that it’s even become hard for my team to keep up. If they take a break too long, the system of doing the work will have changed substantially.

So if you’re anything like me, don’t simply go through the motions, friends. Start making small tests and improvements to your skills with the opportunities in front of you now. Or else you’ll realize later you've wasted years, like I did.

I often get asked about living in BKK so here’s a summary of what I know:

Where to live

Thoughts on where to live:

• Siam / Chit Lom (City-city). That would be like living in the center of town around all the sky rises and traffic congestion. Lively, but busy, noisy and dirty.

• Ekkamai / Thong Lor (Trendy inner-city). Less built up than Siam and Chit Lom areas, there are a lot of cafes and restaurants in this area. It’s also considered to be Bangkok’s Japantown, so you’ll get your fill of Japanese restaurants here. Quieter, but still busy.

• On Nut / Ari (Trendy outer-city). Even less built up than Ekkamai and Thong Lor. These places are up and coming. Ari has lots of cafes and restaurants, but On Nut has more of a living vibe with easy access to a regular shopping mall, cinema and BTS station all in one.

You can check out detailed notes on each neighborhood on Hoodmaps (link). Some of the notes are pretty funny (and true).

You can also read a review post by local startup guy, Oliver Eidel breaking down the costs, pros, cons and views of 3 different condos in BKK here (link).

Other tips

• Stay away from Nana. And probably North-East Asoke also known as Soi Cowboy territory. Unless you love 24-7 red light districts.

• Avoid Airbnb for long stays, unless for a 1-month stay. Rentals are cheaper but with fewer scams. Airbnbs are notorious for overcharging for their own bills, with crazy ‘house rules’ etc.

• Please, walk to the station. Traffic super sucks in Bangkok. I’m not joking when I say you could be caught in a taxi during a traffic jam for 1 to 1.5 hours to cover what could be a 20-minute walking distance. Traffic jams are common and almost every day. Stay within a 10-minute walk of a BTS or MRT (same thing as BTS but a different company) train line and walk.

Where to work

• True Digital Park has a Thai startup vibe with younger folks, but it’s a bit far out, about 35 minutes from Asoke. (link)

• Fabbit Global Gateway. Quiet and low key like a library. Just the way I like it. A lot of Japanese people work here too. It has a cafeteria with a fridge so you can bring your own food. (link)

• The Hive, Thonglor. Hot but lively, more like Chiang Mai digital nomad experience. I haven’t been there more than once. (link)

• Justco, WeWork. I had a great experience with JustCo in Korea. WeWork is pretty similar everywhere, but here it has that corporate startup vibe. I don’t think you’ll find any boot strappers, freelancers, indiehackers etc.

Where to connect

Here are some things to keep in mind about the Bangkok community vs other digital nomad locations:

• Everyone is spread out. Unlike Chiang Mai, Bangkok digital nomads are really spread out and work in all kinds of places. Probably most people work from home, rather than a coworking space.

• Social investment is required to build a network. The people I know who have established friendship groups here tend to have built up the connections slowly. Try to initiate hangouts yourself with other people new to the city because breaking into an existing circle is much harder. I notice even at my own events, it’s basically a bunch of different people’s friendships groups coming together rather than a proper mixing of people.

With that said, here are the most common ways to network with the community out here:

• Meetup.com. t’s the classic. Not scientific, but I think the events here are a bit broader in scope, with more fun outings like hikes or bike-riding trips. But less good for finding professional interest based groups like an SEO meetup (there’s a group for that).

• Facebook groups / events. Digital Nomads Thailand (link), Bangkok Expats (link) and Bangkok Digital Nomads (link).

• WhatsApp groups. The big one is BKK Expats & Digital Nomads (link). But there are niche specific ones floating around. You can ask in the main group and someone will share them with you.

• Newsletters. There’s a free local newsletter, Bangkok Newsletter, which does a great roundup of local events and news (link). Also more generally, check out The Digital Nomad Newsletter (link) which shares info more generally relevant to nomading in Thailand too.

Budgeting

The upper ends on these ranges are based on trying to live an outwardly fancy life:

• Rent 10,000 - 30,000+ THB, for a one-bedroom condo. You’ll pay more in fancier areas like Siam, Ekkamai and Thong Lor. These condos typically have glitzy amenities (literally) like rooftop pools and marble-clad reception areas. On Nut and outer-city areas can be around the 10k-25k mark for a single bed condo. Keep in mind that young university graduates may spend around 6-8k per month in comparison. 3k-ish for a single room.

• Food 300-2000 THB p/day. You can get by cheaply eating low quality street food or cooking a few basics at home. Or, you can eat out at nice restaurants often and get everything by delivery apps.

• Electricity 1,000 - 2,500. If you keep the aircon on all the time in an average one-bedroom condo, it’ll probably cost around 2,000 THB per month.

• Transportation 1,000 - 2,000 THB. One trip via the BTS is about 35 THB. And taking a taxi could average around 250 THB. So you can calculate from there.

• Internet 1,000 - 2,000. There are multiple companies to choose from. 3BB is the cheapest, AFAIK, and offers the best quality for the price.

• Phone credit 350 - 500 THB. You pay more for roaming and bigger data packages, but honestly, I find the minimum already provides an excessive amount of data and minutes.

These will be the most common expenses.

Language Learning

I tried learning with a language school (which I won’t name), but my experience wasn’t great, so I ended up learning with tutors and a course.

• Duke Language School is probably the most famous here and the one I see most recommended on Facebook groups. They seem the most legit, too. (link)

• RapidThai is a course to teach you Thai quickly with a pre-made mnemonic system. I took their course on learning to read it is great. (link)

The textbooks here don’t seem to be that good when I compare them to Korean language learning textbooks. But I have found one reasonable one which is COMPLETE THAI by Teach Yourself. Apparently it will take you to B1 (mid) level by completion.

Transportation

I’d say you’ll end up using the BTS/MRT 75% of the time and taxis 25%.

• BTS/MRT. You need to bring your passport to the station and request a BTS card. The MRT is just another BTS-type service, but run by a different company, so you’ll need a separate card for it. If you want to get points for all transactions so you can use them for free stuff or discounts, you can download the Rabbit points app and attach it with your BTS card.

• Taxis. Get Grab! And Bolt, and Robinhood and LINEMAN apps. These are the local Uber and delivery providers, with Grab being the main one. You’ll need multiple because sometimes Grab is too busy so you can use another to book a ride or delivery instead.

• Tuktuks. Honestly these things just seem to be tourist traps. Avoid unless you’re looking for a tourist experience.

• Motorbike taxis. Dangerous but convenient, so be wary. The problem is that often you aren’t provided a helmet, and if you get one it’s going to be nasty inside. Sometimes you can see the riders themselves are a bit drunk late at night, so you’re rolling the dice with those guys.

Eating and going out

I’m not a foodie and don’t drink so my suggestions are almost useless. This is the best I’ve got:

• Emquartier Starbucks has an awesome view.

• Icon Siam is a super flashy mall and impresses visitors.

• The Grand Palace is very nice. Definitely worth going to as a tourist.

Banking

There’s mixed info about whether you need a work permit and certain visas to get a local bank account. I did mine with a study visa with the help of a school. Honestly, seems like it depends on the branch as some people have done it with tourist visas. You’ll need to try by yourself.

Typical scams

Thais are great, and honestly, I’ve had little trouble with scams or low-integrity people, so I wouldn’t say you need to be overly cautious. But unfortunately, scams are more prevalent around tourist areas.

• Airport taxi. Sometimes the taxis quote you a crazy amount to drive back from the airport, over 400 THB+ (obviously depends on how far you live away). Be wary. And the meter can sometimes be switched to night mode, so you get extra charged.

• Tuktuk drivers saying “it’s closed dawg.” They’ll tell you a main attraction is closed and offer instead to take you on a ride to see their friends’ places instead.

• Bus-friend scams. Someone at the ticket sales booth will sell you a ticket for a bus operated by their friend, which does not go directly to the destination. Instead, you’ll get on a smaller van bus and stop everywhere on the way so that the driver can collect more fees from random passengers along the way.

• Tailors. Some tailors show you one fabric but deliver you another once it’s done.

• Fake watches. Someone will come up to you in the street around Nana and try to hustle you a watch (it’s fake). (link)

Other tips

Suggestions that didn’t fit anywhere else.

• PM 2.5 filter + mask. Essential. The air is bad around March-April at the start of the year and again at the beginning of winter.

• You don’t have to wai everybody. It seems Thais don’t wai a waiter or waitress, for example, but I see some foreigners do it when they first arrive. You can wai elders/important people, though it’s not expected of foreigners. I save it for when dealing with formal occasions.

“I know I need to make content, have an email list, etc. But I’m unsure of the big picture and how everything fits together.”

I’m not an expert, but I’ve spent possibly 50+ hours nerding out on this topic.

These are my notes and I hope they help if you’re an online coach or indie founder figuring out your content strategy.

Every platform has it’s strengths and weaknesses. So use different channels for each part of the funnel.

Creator content funnels seem to involve 3-4 major moving parts that help you get found, capture emails and direct your following to build trust or sell stuff.

These channels are:

• 👀 Discovery channels. Platforms like Instagram, LinkedIn, Twitter, or even Facebook. Places you can post short, quick content and get free organic visibility without an audience, that builds up over time.

• 🧲 Lead magnets/email capture. Helps you build your email list by offering giveaways on your discovery channels or suggesting people subscribe on your other channels.

• ❤️ Retention channels. These are things like Skool communities, Facebook groups and email lists. Jay Clouse from Creator Science mentions they’re for building relationships at scale (link). Used to direct attention to products or other content, too.

• 🧠 Long-form channels. YouTube, blogs and podcasts. These channels are used to nurture your followers to know, like and trust you.

You don’t need all of them. Some people achieve all 4 functions with only Instagram or Twitter.

It’s commonly suggested that fewer channels are better than more. Running each part is a separate skill, and audiences compound, so spreading yourself thin across multiple platforms for each part is unnecessary.

Examples

Here are some examples demonstrating different creator content funnels.

Vanessa Lau

Vanessa used Instagram in 2018 and added YouTube in 2019, teaching coaches how to use both channels effectively.

• Posts frequently originally on Instagram, used reels to attract followers and slideshow posts to nurture them.

• Used captions on her posts to direct people to webinars like How to Turn Followers Into Customers or other content.

• Probably used a ManyChat automation to auto-dm reply people who used keywords in the comments. Learn how to do that here (link).

• Previously she seemed to rely a lot on DMs and webinars to promote her group coaching program, which was very highly rated.

Justin Welsh

Justin is famous for popularizing the content-marketing based solopreneur movement overall using LinkedIn as his main discovery channel.

• Posts daily on LinkedIn. Engages daily in the comments to improve performance.

• Started Twitter later to diversify and leveraged his LinkedIn audience to build his following.

• Plugs his newsletter on high performing posts in the comments or directly in the post.

• Broadcasts 4-minute articles to his newsletter every Saturday, also posting them on his blog.

• At the end of each email, he gently promotes his courses.

You can learn exactly how he does content for his solopreneurship in his course The Content OS (link) (affiliate link).

To make starting easier, get started with fewer channels, not more.

I agree with that it is better to keep things simple rather than spread yourself thin. Pick one platform. Most personal branding experts tend to say you should go where you already spend time. It’s also probably right to pick channels that your target audience uses, too. If you can.

And if you wanted to add one more piece to the funnel, email seems to be the most important according to teachers like Nicholas Cole.1

For me, Substack gives me some level of discoverability via it’s it’s Twitter clone, Notes, and I get to use it’s native newsletter function for retention too. I then use my YouTube channel to build that trust in my profession as well as blog here, too.

What’s your plan?

Lead generation Facebook advertising freelancers are probably going to get whacked by AI in the long term.

But exactly how close is AI to taking our jobs away in February 2025?

To see the state of play for AI in Meta ads, we can break it down into four functions:

1. Copywriting

2. Creative Design

3. Media Buying

4. Strategy

In the future, we might have all four mixed into one function plus the functionality of auto-created landing pages. But let’s review each one at a time.

1. Copywriting

Most people have probably played around with using ChatGPT to generate text for something.

And probably also noticed that the generic output sucks.

But at the risk of being a bro, I have to say, 'You didn’t prompt it right.' AI-generated copy can yield great results for immediate use or with some editing.

Here are 6 example levels of copywriting prompting:

1. Generic instructions, “Write an ad for XYZ company”

2. Specific instructions, “to ABC audience, promoting features 1,23”

3. Specific with examples, “… for example {insert from other ads you’ve written}

4. Step by step with examples,

   1. “first, understand this business such that you could write Meta ads copy when promoted. {Business details}.”

   2. “next understand this product and market such that you could write Meta ads copy when prompted. {Product & target details}

   3. “now write an ad for XYZ company… to audience … for example…”

5. Step by step with modular examples, "Write an ad for XYZ company with this framework Headline, Subhead, Product Intro, Product Details, Social Proof and CTA. Here are 3 examples for each element…”

6. Step by step with modular examples AND tweaking, “Now, evaluate the copy from the POV of a {style reference} and re-write based on this.

The output is completely different.

Here’s example ad copy for a Joe’s pizza shop using generic instructions which is where most people start.

Generic prompt (no no tweaking or editing):

Write Facebook ad body copy to promote a pizza shop called "Joe's Pizza".

🔥 Craving the Perfect Slice? 🍕

At Joe’s, we don’t just make pizza—we craft mouthwatering, oven-fresh perfection. Whether you love a classic cheese slice, a loaded meat lover’s pie, or something extra cheesy and gooey, we’ve got you covered.

✅ Hand-tossed, fresh dough daily

✅ Premium, melty cheeses & bold flavors

✅ Baked to crispy, golden perfection

📍 Swing by Joe’s today or order online for pickup & delivery. Because great pizza = great moments.

👉 Tap below to order now! 🍕🚀

Using a level 5 prompt: (no examples, no tweaking or editing):

• Copied and pasted the info from the Joe’s Pizza website

• Copied and pasted probably 50-100 Google reviews + Google maps basic map data

Write Facebook ads body copy for Joe's with this framework call out, product intro, Product Details, Social Proof and CTA targeting locals for the cheese slice with a 50% off discount coupon promotion. For the call out, pick out a quote sentence that could be used as a headline.

🔥 “Joe’s is worth every bit of the hype and the block-long line!” 🍕🔥

Locals know: Nothing beats a classic New York cheese slice from Joe’s Pizza! And for a limited time, we’re treating our NYC neighbors to 50% OFF your next slice!

🍕 What Makes Our Cheese Slice Legendary?

✔️ Crispy, foldable thin crust – the true NYC way

✔️ Perfectly balanced sauce-to-cheese ratio

✔️ Freshly made, straight from the oven

But don’t just take our word for it…

⭐ "The best slice I’ve ever had—crispy, cheesy, and pure perfection!"

⭐ "This is what NYC pizza should taste like. Period."

📍 Exclusive for Locals! Grab your 50% OFF coupon now and enjoy the city’s best slice at half the price!

👉 Claim your discount now! 🍕🎟️

After 1 tweaking prompt (no extra examples):

Critically analyze this copy to see whether it meets the requirements of being non-cheesy, non-try hard from the point of view of a smug Redditor. For example, consider the analogies, tone and wording used. Then, rewrite it based on this analysis.

“It’s just a damn good slice. Nothing fancy, no gimmicks.”

You already know the deal. It’s a proper New York cheese slice, made the way it should be. No overloaded toppings, no weird crusts, no nonsense—just the classic, done right.

And for once, we’re knocking 50% off your next slice—because why not?

Not forever. Just for now. Grab it while it’s here.

👉 Claim your 50% off now.

Not bad. Especially given that I used a basic framework and there’s no editing yet.

Meta is experimenting here too.

When you add your own copy, in some accounts, it will show you it’s own AI generated versions that you can test against your own.

And they’re scarily decent.

Headlines, too.

Given that Meta can see all the ads written on their platform and whether they’re working, I wouldn’t be surprised if this means advertisers will eventually just write one version of the copy and then generate the rest on platform.

2. Creative

Images

AI image creatives are far from easily usable, but there is a lot of activity here.

Firstly, Meta is playing with generating creatives itself (see below). They’re pretty cruddy but I think it could work better in eCommerce advertising where people and location context is so crucial to communicating what the ad’s about.

They allow you to generate backgrounds (presumedly for just eCommerce items) or the entire image. Then add and edit the text overlayed ontop:

Other applications are springing up like Design.com, AdCreative.ai and Predis.ai, but while they’re working hard - so far they’re basically an image generator with a Canva subscription.

Take Predis for example.

It is super cool that it does the copy, images and formats all in one using an AI wizard approach. But the results are pretty underwhelming:

The most likely and practical way I’ve seen to make this work is to build your own AI ad generator with a mashup of ChatGPT, Figma and Midjourney:

Video

Surprisingly, it’s the video side of ad creatives that looks like it will get off the ground first.

Creatify has a wizard AI ad generation experience where you put in your company, product and audience details. Then it will generate a script (or you can add your own) with an avatar of your choice.

Here’s an example video ad I made using General Assembly’s digital marketing short course as the sample data:

What’s cool is that you can add your own creative assets to run in the background:

I’m looking to explore this more because with the right script, I think I can make this work.

(Let me know if you’d me to write like a follow up piece about using this tool in the comments).

3. Media buying

Media buying on Facebook ads can be broken down to:

• Targeting

• Budgeting

• Placements

And truthfully, the Meta media buying process has become more and more simplified overtime.

Overall, Meta does offer simplified Tailored campaigns which set up everything for you though not recommend if you consider how they’ve been juicing people using the boosted posts feature.

Budget

The AI automation is already here and prevalent.

A large group of people use Advantage campaign budget which automatically allocates ad spend to targeted groups within your setup that should give you the best performance.

Targeting

For targeting, it’s mostly become a non-consideration too as most people put the bare minimum targeting settings on Facebook ads, filtering just by gender, language, location and age.

A newish feature has been the Advantage+ audience targeting option which relies more on Facebook’s internal targeting parameters, beyond the settings you put in place, to find buyers.

Many get mixed results.

Personally I use it sparingly as the results can be inconsistent. And sometimes you need to set a firm line not to target a certain group of people.

Placements

Placements settings on Meta tell the platform where you want the ads to be shown. For example, on Instagram or Facebook, feeds only or stories and reels.

Facebook has an automation option here called Advantage+ placements which is decent except in certain cases where you can accidentally optimize for poor quality responses.

For example, Advantage+ includes placements on third-party apps and websites:

It’s not consistent whether another website will show your images, headlines or body copy properly (or even at all) which makes it really easy for people to misunderstand your ad.

In one case, I think using the apply now call to action on an add plus the Advantage+ placements may have put my ads onto a job website so people responding thought they were applying for a job. Problem was that the ad was for a paid coaching program!

4. Strategy

Strategy in Meta ad buying includes a few parts:

• Understanding a client’s goals, abilities and resources

• Choosing customer targets

• Researching them to understand desires, fears and interests

• Formulating offers

• Planning funnels

Then mashing them together to create an advertising plan that is effective, profitable and workable for a client.

Predictably, there’s no meaningful progress on this part of the job yet. And I think that’s for a 2 reasons:

• There’s a barrier between the meat-space and cyber-space. Someone has to get that client data on their situation into a machine. AI isn’t quite ready to quiz a client all on their own.

• Inconsistent training materials. Most of Reddit can’t agree on how to run an ad account. So how can AI be trained on that? Even Meta’s own training materials go out of date quickly due to updates in the UI, new features, and the regulatory environment, so they aren’t a reliable source of training material for an AI.

For the adventurous few, I suppose you could try DIY your own part-solution using Chat GPT Projects.

I made my own by adding Meta’s official training materials as PDFs so that I can quiz them on my approach to certain campaign structures.

While it is useful for media buying setup feedback, ultimately it’s not even close to replacing a human with experience and judgment.

While AI in February 2025 isn’t close to replacing an experienced Meta media buyer yet, it is making fast progress.

Given the automation in the media component of the job, I think it’s smart to hope your freelancer or agency is leveraging AI to increase their productivity and spend more time making creative, persuasive ads that rock.

Well, that sounds like content marketing.

This was the answer to a question I asked myself, “What would be the best marketing skill of all to develop?” as a marketing freelancer (now focused on Meta Lead Advertising) looking to bring in work over time.

I had hoped this could become a new service offering that I could master for myself and my clients.

And content marketing seems relatively simple, reading the advice of high performers like Justin Welsh and Nicolas Cole (I’m a fan of both, btw). Just pump out a high volume of work, see what works, refine and repeat.

Taking Action

Following this, I made 40+ podcast episodes and 68 long-form articles over 18 months, which were promoted across Facebook, Twitter, LinkedIn, Indiehackers, Reddit, Medium, etc.

My little podcast was listened to thousands of times and articles the same.

Sitting back, I assumed it would eventually pay off with even a trickle of emails, replies, comments etc.

The Results

But an audience didn’t materialize.

People read and listened but they didn’t respond with likes or comments.

Despite spending hours planning, writing, recording and editing pieces, they weren’t often shared beyond a few hit podcast episodes.

I think for the 1.5 years of my podcasting efforts there was single inbound enquiry.

Content Promotion

Naturally, beyond improving the content, I tried to fix this by pushing harder with promotion. I pumped my LinkedIn, Twitter feeds, Slack posts, groups, etc with shared links.

But it didn’t make a difference.

While the actual content journey itself was fun, embarrassingly, in reality very few people wanted to be part of my ‘audience’.

Thankfully but sadly, it’s not just been my experience it seems. Content overally is a tough game:

• A 2022 survey by the Content Marketing Institute itself found that “72% of marketing professionals today describe their content marketing campaigns as falling between flat-out ineffective and only mildly effective.” (Source)

• In fact, per eMarketer/Goldman Sachs in 2023, only 4% of content creators make more than 100k in revenue (Source). But to be fair, that’s a pretty high bar… 🤔

Going Beyond Transactional Content

Then my friend shared with me a point of view that taught me where I went wrong:

It is better to have 500 people who know (and like you) than 50,000 people who follow you for memes.

That was it. Content production by itself is just motion.

To be effective, content has to go beyond an information transaction and instead build that know, like and trust factor.

Content marketing in this instance is not about volume of articles or reach. It’s a digitsed networking engine.

This frame change means focusing less on output and more on creating pieces that are relevant to a target reader but also help build a strong connection to myself.

That means including myself into the work, not just pumping out anything I think could be useful.

For example:

• I created a lot of marketing interview podcast episodes with the hope that guests would share the work, too (News: They usually didn’t). I should have made solo episodes on my topic of expertise.

• I needed to focus on topics that naturally feature yours truly and my topics of expertise. Rather than publish so broadly, I.e. general marketing content.

• Lean more into connection-building formats like video over text. (Note: This is one reason I made a YouTube channel).

• Included more photo or visual representations of myself. In action 😏, potentially to connect the reader with the writer.

• Be honest with myself about when I’m publishing as a hobby (like this article) and when there is a commercial purpose to avoid wasting time.

Thinking aloud, this is probably why:

• We remember more personable, emotive writers — even if we don’t remember what we learned from them.

• Nobody honestly remembers the last PDF lead magnet they downloaded.

• I can’t name a single HubSpot blog author 😂 despite them successfully dominating half the web...

Especially now, there is so much AI-generated content that even recognizing an author can be hard.

And reading around after the fact, I find the view is well understood for podcasts, as explained by Jason Cercone.

I think content is still a powerful marketing tactic but I’ve learned that it has to be done in the right way:

Focused on connecting oneself with the reader through relevant topics with a clear personal connection.

P.S. If you to sidestep the content marketing journey and get directly to the email signups and leads part with Meta ads, feel free to send me a message.