As a developer, I love to build.
Planning it out, picking the stack, getting into flow state, watching something go from nothing to a working product.
That part is fun. And with AI tools in the mix, that part is faster than ever.
But here's what I've been facing in reality: shipping the code is maybe 30% of the work. The other 70% is everything that comes after.
I want to walk through what that looked like in practice for my recent project. Not quite a tutorial or success story. Just an honest look at the messy middle.
The Build
A few months ago I built FoFoClip, a tool that takes a podcast episode and turns it into polished, ready-to-post video clips. You upload your audio, pick a style, and get clips delivered to your inbox. Supports 6+ languages.
The tech stack is my usual go-to: Vue 3, Quasar Framework, Firebase (Hosting, Auth, Cloud Functions, Storage, Tasks), with AI handling the heavy lifting on the backend for transcription, clip selection, and caption rendering.
I used AI pair programming for the entire build. Claude for architecture decisions and complex logic, Claude Code for fast iteration, and a workflow where I'd break things into focused prompts rather than trying to generate everything at once. If you've read my earlier posts, you know I'm a big believer in this approach.
The build took about a solid week of evening work. Prompting, testing, revising, prompting again. Security reviews. Edge case handling. Making sure the clip styles actually looked good across different types of content.
And then it was done and deployed! Real product offering, real Stripe integration, real pricing.
I felt great. I had built something real. Until haha...
The Silence
Here's the thing about launching a product as a solo developer: unless you built some hype or awareness... nobody is waiting for it.
There's no Product Hunt launch with 500 upvotes. No Twitter thread going viral, etc. I push to production, maybe tell a few people, and then... quiet.
I told my mom. I posted a bit on social media. And then I kind of waited for something to happen.
Nothing happened.
So I did what a lot of us do when things aren't moving. I went and worked on other projects. I had plenty of them. It felt productive. It felt like I was still shipping. But really, I was avoiding the harder work that FoFoClip needed.
The YouTube Ads Experiment
Eventually I came back to FoFoClip and decided to throw some money at it. I made a 30-second video ad, set up a YouTube campaign targeting podcast-related search terms, and put $50 behind it.
The results: 59 website visits. Zero conversions.
While I was proud some people were clicking (though honestly, they may have been ghost clicks...) not one person signed up. Not one person bought a DANG THING.
Now, $50 on YouTube ads is barely a test... but zero conversions out of 59 visits still stings. It made me look harder at the landing page, the messaging, the trust signals, all the stuff that has nothing to do with code quality.
A couple things I found:
- Perhaps the pricing was too high?
- Perhaps the brand was confusing. The product is "fofoclip," the domain is clip.fofo.dev, support emails come from sweetpapatechnologies.com. That's a lot of "who are these people?"
- I had no analytics that actually worked. My cookie consent banner was so privacy-friendly that most visitors opted out and never got tracked. I literally could not see my own traffic.
None of this is a code problem. This was even bigger. It was a marketing problem!
Google Won't Index My Site
After the ads experiment, I noticed that searching for my site on Google returned nothing. I assumed it was a technical issue. Spent time making sure everything was statically rendered at build time. Checked my robots.txt. Checked my HTTP headers. Checked for noindex meta tags. Everything was clean.
Then I looked at Google Search Console and saw this status:
"Crawled - currently not indexed"
So... basically Google successfully crawled my site. It confirmed that indexing was allowed. It fetched the page without errors. And then it chose NOT to index it.
The reason likely: trust. My site has zero backlinks. No referring pages. No external mentions anywhere on the internet. Google looked at clip.fofo.dev and essentially said: "We see you. We just don't have a reason to care yet."
This was a humbling realization, because I have never had that happen before... You can have perfect headers, perfect SSG, perfect robots.txt, and Google will still shrug if you haven't done the non-technical work of building credibility on the web.
What I'm Learning
I don't have a framework for this or big tips. I'm still in the middle of it and still learning. But here's what's becoming clear:
Building can become the comfort zone. Developers default to solving problems with code because that's what we know. But sometimes the issue is much more about UX and non-architectural
Marketing is a skill, not a chore. I used to think of marketing as something you tack on after the real work is done. It's not, it is very much a big and standalone thing. Understanding why someone would trust a new product enough to pay for it, that's a design problem just as interesting as any architecture decision.
The build never really ends, but it needs to share bandwidth. I still have bugs to fix and features to add. But if I spend all my energy on code and none on outreach, content, and trust-building, I'll have a really well-engineered product that nobody uses.
Talk about what you're building. This is something I'm getting better at. Blog posts, Bluesky updates, dev.to articles. Every piece of content you put out there is a signal to both humans and search engines that you exist.
A.I. is super helpful for beyond just coding. I was able to make ad content, promotional material, and get all sorts of questions I had about marketing answered.
For a bootstrapped solo project, content might be the most cost-effective marketing channel there is.
Where I'm At Now
FoFoClip is still live at clip.fofo.dev. I'm still working on it.
I also have a bunch of other SPT projects in various stages. The temptation to go heads-down on those is real. But I know from experience that building another thing won't solve the problem that this thing has.
The code was the easy part. Of course there were challenges and headaches along the way, but getting people to find my service, trust it, and pay for it... that's the actual product work. And I'm just getting started on it.
It will be interesting to see if this app and service gets any use, and serves a purpose for someone.
That is what makes coding fun, is seeing the product come to life.
