1
0 Comments

Unexpected Lessons From 30 Daily Users: How ChatGPT Became My Top Traffic Source

Unexpected Lessons From 30 Daily Users
How ChatGPT Quietly Became My Top Traffic Source

For the last few weeks, I’ve been building a small side project in public.

No launch post.
No ads.
No growth hacks.

Yet for the past 20 days, the product has been getting a steady 20–30 unique users every single day.
At first glance, that doesn’t sound impressive. But once I dug into the data, the story behind those users turned out to be far more meaningful than the numbers themselves.

This post is about what those early users taught me.

The Product (Quick Context)
The product is a simple utility.
It helps people check whether an image looks AI-generated or manipulated.

There’s no signup.
No onboarding.
No pricing page.

You upload an image, get a result, and leave.

I originally built it to solve my own problem around image trust. I didn’t know if anyone else would care. Turns out, some people do.

Consistency Was the First Real Signal

The most important signal wasn’t growth.
It was consistency.

For nearly three weeks:

  • 20–30 unique users every day

  • Visitors from multiple countries

  • Zero paid marketing

  • No spikes.

  • No viral moments.

  • No launches.

Just quiet, repeated usage.

That told me something important:

This wasn’t noise.
People were actively looking for this solution.

Consistency beat excitement.
The Metric I Did Not Expect
When I opened analytics, I assumed Google would be the main traffic source.

It wasn’t.
Almost half of all users (~48%) came from ChatGPT.
That stopped me.

People were asking questions like:

  • “How can I tell if an image is AI-generated?”
  • “Is this photo fake?”
  • “How do I verify image authenticity?”

And instead of sending them to a blog post or a forum, ChatGPT was pointing them directly to my product.

That’s when it clicked:

AI assistants aren’t just tools anymore.
They’re becoming discovery engines.

  • This wasn’t SEO.
  • This wasn’t social.
  • This was answer-driven traffic.

These Users Had Real Intent

Traffic alone doesn’t mean much. Behavior does.

Out of roughly 511 users:

  • 126 uploaded an image
  • That’s about a 25% activation rate
  • No signup
  • No hand-holding

For an early, cold, utility product, that’s a strong signal.

These users weren’t browsing.
They weren’t curious clicks.

They arrived with a problem and tried to solve it.

Some patterns stood out:
Google users had the highest engagement
ChatGPT users made decisions quickly
Hacker News users were fewer, but explored deeply
Different sources. Same intent.

A Reality Check: This Is Not a Retention Product

Most users upload once and leave.

They don’t come back the next day.
They don’t form a habit.
And that’s okay.

This isn’t a daily-use app.
It’s a moment-of-need tool.
The real goal isn’t retention.
It’s recall.

“Next time I need to verify an image, I know where to go.”

Understanding that saved me from chasing the wrong metrics.

on December 25, 2025
Trending on Indie Hackers
I built an AI that turns an idea into a live business in under 10 minutes. Here’s what 1,000 launches taught me User Avatar 76 comments I built a web-based vector editor from scratch and integrated an AI Agent. Need just ONE beta tester! User Avatar 70 comments Building Noodle, a keyboard-first REST client for the terminal User Avatar 31 comments "Looks Good to Me" Is Quietly Killing Your Feedback Loop User Avatar 28 comments Launched 580 landing pages in 1 week. Solo. No team. User Avatar 18 comments I built a competitor monitor for indie founders User Avatar 14 comments