It’s quite chilly here in Medellín.
Jordan, Youssef and I came here for a locked-in period in April and May. Waking up at 5am, going to sleep at 10pm. Eat, work out, build. Nothing else.
In the past month I killed Good XP, burned a billion tokens, and learned one crucial fact about consumer apps that I should have known from day one.
Let me tell you the news.
Good XP promised weight loss.
Follow the plan, eat less, move more, you will lose weight. Calories In, Calories Out is the only thing that matters, and I am living proof of that. Three times over.
But the promise required the user to eat less than they were used to, hold their hunger with willpower alone, and wait weeks before seeing the first result.
I was asking faith from my users.
And even if they stayed faithful, a hundred other things could get in the way. Stress. Travel. A bad week. The app had no answer for any of them.
I was shipping a consumer product and treating it like a therapy program. That doesn’t work.
On one of our walks through in Medellín, I was venting to Jordan and Youssef about the problem. Weak retention. A value prop that only paid out in month two.
Jordan casually mentioned:
“Thiago, you’re good at tracking. You’re organized and always on top of everything. You track things pretty well.”
Tracking… Then it clicked.
Weight loss worked for me because I always tracked what I ate. I found strategies to keep tracking even when my reptilian brain ran for cheap dopamine. I learned to estimate a plate of food after hundreds, maybe thousands of reps of tracking what I ate.
Tracking…
It was tracking all along. Not weight loss.
Tracking was the skill that unlocked weight loss for me. And dozens of other goals before that.
I’d built an app about the outcome. I should have built it about the behavior that produces the outcome.
Here is the lesson I paid a year of my life to learn:
A consumer app has to deliver value in the first 30 seconds. Not 30 days. Not 4 weeks. Thirty seconds.
The user downloads. Opens the app. Something has to click immediately. A number they didn’t know. A view of themselves they couldn’t see before. A tiny win that makes them come back tomorrow.
Good XP’s value was locked behind a month of discipline. That’s a great newsletter. Not a great app.
Weight loss is lagging. Tracking is instant. You scan a plate, you see a number, you feel informed. You did something. The value delivered itself before you even finished breakfast.
I was selling the mountain. I should have been selling the first step.
A week ago I rebranded the whole thing. Good XP is dead. clrr 🥭, the AI calorie tracker, took its place. (Pronounced “clear.” btw)
And may I be completely honest: it is much more aligned to who I am as a founder.
I’m not pretending to be a weight loss coach. I don’t want to be one. I want to be the person who builds the sharpest, cleanest tracking tool on the market, for the person who already understands that tracking is the unlock.
Same underlying math. Much smaller ask. A much more clear 🥭 product.
I spent a weekend burning every token I could find. Rebranded the codebase, the website, the language, all of it. Ran out on Cursor and Claude Code Max (fucking limits). Moved to Antigravity, TRAE, Copilot, Windsurf, Kiro. Ran dry on those too. I have now tried every coding agent under the sun.
The shipping part was easy. The hard part was admitting I’d been pointing at the wrong target.
Here’s the new problem.
Calorie tracking is a blood-red ocean. MyFitnessPal. Lose It. Cal AI. A hundred others. Every one of them has more users, more data, and more budget than I do.
I picked a category where time-to-value is real, and I’m late to it.
So now I get to learn the other lesson I’ve been avoiding: how to find a wedge in a saturated market without pretending the market is empty.
That’s the next piece.
— Thiago Ricieri
X.com @thiagobuilds
Instagram @thgvr or @thiagobuilds
Threads @thgvr or @thiagobuilds
LinkedIn @thiagoricieri
Substack Making of a Maker written by @thgvr
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.