Less.
Better.
Full stop.

That number isn't a feature. It's an indictment. When a platform hosts 2.5 billion products, "search" becomes a lottery, "best seller" is a paid placement, and "customers also bought" is a targeting algorithm with a sales quota. The catalog isn't curated — it's uploaded.

Every "best of" list you've read in the last decade was probably written by someone who tested none of the products. The affiliate link came first; the review was built around it. The comparison site exists to collect your click, not to help you decide. The sponsored badge is there if you look, in a font size calibrated so you won't.

This is the retail internet as it currently exists: infinite shelf space, zero editorial spine, and a financial model that rewards whoever bids the most — not whoever makes the best thing.

Curation is a skill. It requires knowing what a thing is supposed to do, forming an opinion about how well it does that thing relative to everything else at the same price, and being willing to exclude the 47 options that are almost as good. That last part — the exclusion — is where most "curated" collections fail. Saying yes is easy. Saying no is the product.

The rarest thing in retail right now isn't a great product. It's someone willing to commit to one and stand behind it.

We're not anti-Amazon. We're anti-abdication. Using a 2.5 billion product marketplace to find one kitchen knife is a task that no normal person should have to do, and that no algorithm currently does well. The signal-to-noise ratio is too broken. The incentives are too misaligned. The reviews are too gamed.

We think most people, most of the time, don't actually want more choice. They want the confidence that someone already figured it out. They want to know: of everything available right now, at a price that isn't absurd, what is the one I should own?

That's the question we're answering. That's the catalog we're building — slowly, with a high bar for what makes it in, and no shame about keeping it small.