Everyone keeps telling me to use Google Lens. Here's what I learned building Estimatik.
Every time I mention Estimatik, I hear the same response:
"Why not just use Google Lens?"
Fair point. Google Lens is impressive at what it does.
Point your camera at a jacket, lamp, sneaker, or random garage sale item, and it can often tell you what it is.
But while building Estimatik, I realized something important:
Identification and valuation are two completely different problems.
Knowing an item is a "Nike jacket" doesn't answer:
Is it worth $15 or $90?
What did similar ones actually sell for on eBay?
Should I buy it right now or walk away?
That gap is where most thrift hunters and resellers lose time. They identify the object… then start manual research across sold listings, condition comparisons, model names, and guesswork.
Last week someone scanned a $3 Goodwill find — turned out it was worth $85–120 based on 24 recent eBay sold listings. That's the kind of answer Google Lens doesn't give you.
That gap is exactly why I built Estimatik:
Take a photo → get an estimated resale value based on real eBay sold prices.
Building this has taught me something useful:
A lot of products look "already solved" from the outside. But once you study the real workflow, you often find the painful step nobody solved properly.
Curious if other founders here have experienced the same thing:
Have users ever told you your product already existed… only for you to discover the real problem was somewhere deeper?
Recognition is free, valuation is the product.” That’s a one-line pitch. The landing page drop-off problem might solve itself if the first thing visitors see is a live demo scan, not an explanation of what the tool does. Let the output speak. The $3 to $85 Goodwill example is your best marketing asset.
That's exactly the angle "Should you buy it? Find out before you pay." The real-world flip examples are already front and center. Appreciate the validation.
The gap between "identifying" an item and "valuing" it is exactly where resellers lose their margins, Messaoud. You've hit on a brilliant point—most people mistake a high-level feature like Google Lens for a complete workflow, but the real money is made in the specific data layer you're building.
I’m currently running a project in Tokyo (Tokyo Lore) that highlights high-utility tools designed to solve these exact types of "hidden" workflow problems. Since you've successfully identified a deep pain point that the tech giants have overlooked, your project would be a standout entry for our current round.
A lot of “already solved” markets are only solved for the first step, not the valuable one.
Lens gives recognition, you’re competing on decision-making.
That's the cleanest way to frame it. Recognition is free and everywhere. The valuation layer is where the actual decision happens, buy or pass. That's the gap.
This is a very creative tool, and if this direction is developed well, it's easy to integrate it into real life (for example, helping people shop better).
I used your tool to analyze a toy; it was very fast, and the interface was clear and easy to understand.
It's normal for users to leave tool-type apps immediately after use; it would be better to find scenarios where users get used to opening them.
I think the only problem is that most users probably only have Google services, and switching to other apps requires some guidance.
A mobile app version would be even more convenient.
Thanks Heng_T glad the scan was fast and clear, that's exactly the experience I'm aiming for.
You're spot on about the habit loop challenge. Utility tools have a natural "use and leave" pattern. I'm exploring ways to bring users back at the right moment — like when they're heading to a garage sale or listing items on eBay. The trigger has to match a real-life situation, not just a notification.
On mobile the web app is already fully responsive, so it works on any phone browser without installing anything. Keeping it web-first lets me iterate faster as a solo founder. But good to know the demand is there.
Thanks for testing it out and for the thoughtful feedback!
This is super relatable — getting traffic early on is way harder than people expect.
One thing I’ve been noticing though: sometimes it’s not just about traffic, but what happens after people land.
Even small UX things (like unclear value in the first few seconds or confusing structure) can make early traffic feel like “no traffic” because people bounce quickly.
Curious — have you looked at how users behave on the page? Like where they drop off or what they actually click?
Happy to take a look if you want another pair of eyes.
https://estimatik.com/
Thanks for that and fair point. Traffic without retention is just noise.
I've looked at the basics: bounce rate is high but session duration on the scan page is decent, which tells me people who actually try the tool engage with it. The drop-off happens earlier, on the landing page itself.
My current read: the value prop is clear enough, but the friction before the first scan is one step too many. Working on that now.
I'd take you up on the second pair of eyes, what would you need from me?