What we've actually found, written up honestly.
Real research from real client work, our own tools, and the occasional strong opinion. Published under our own name. Not marketing copy dressed up as insight.
How to Optimize Your Amazon Listing to Rank on Alexa for Shopping
Adapting listings for AI shopping recommendations, beyond traditional keyword ranking.
Read on amzinsightly.comCPG Brands Are Optimizing for Keywords. Amazon's AI Is Optimizing for Intent.
Why the shift from keyword-centric to intent-driven product content matters now.
Read on amzinsightly.comHow we estimate real search volume for the 76,000+ keywords Amazon won't give you a number for
Amazon's Search Query Performance (SQP) report gives you real search volume, but only for terms your brand already gets impressions on, typically a few hundred to a thousand terms. The Top Search Terms report covers the whole category, tens of thousands of terms, but publishes rank, never a raw volume number. Most sellers treat that gap as unsolvable and stop looking past their own SQP export.
We didn't. We matched the terms that appear in both reports (113 real matches, on one real July 2026 account) and fit a power-law curve, volume as a function of rank, through that overlap. The fit held: R-squared of 0.974. Applying that curve to every ranked term outside the brand's own SQP data gives a real, honest estimate of category-wide demand the brand has never touched.
The honest limits, stated plainly rather than rounded away: the formula only holds inside its own calibration's rank range, anything past ranks 1,417 to 1,211,481 on this dataset is extrapolation and gets flagged low-confidence, not shown with false precision. And on the terms we could check directly, the estimate ran a median 12%, mean 15.7% off the real number. That's a genuinely useful estimate for planning, not a replacement for real SQP data where you have it.
