I’m building a tool to detect marketplace search trends before they become obvious

I’ve been building a small marketplace trend research tool for Amazon sellers, and this is one of the clearest examples I’ve found so far.

The keyword “tubes for river floating” moved from a search rank of roughly 1.27 million to 19,799 over 12 weeks.

Since a lower rank means stronger search demand, that represents a 98.4% improvement.

What interested me wasn’t only the size of the move. It was the structure behind it:

• 11 out of 11 comparable weekly periods improved

• 100% historical coverage

• 100% consistency

• Moderate volatility

• Continued positive movement in the most recent weeks

The tool scored it 97/100 — Breakout.

This is the problem I originally wanted to solve for myself: search volume can show what is already popular, but it doesn’t always show whether a keyword is steadily gaining momentum or merely experiencing a temporary spike.

So I started analyzing weekly search-rank history using signals such as:

• Total rank improvement

• Week-to-week consistency

• Recent acceleration or slowdown

• Volatility

• Amount of supporting historical data

The goal is to surface marketplace demand while a trend is still developing, before competition and advertising costs fully catch up.

I’m now turning the internal tool into a small SaaS called Uptrend Hunter.

I’d be interested to hear how other founders would approach the scoring. Would you prioritize consistency, acceleration, total improvement, or volatility when deciding whether a trend is real?