Why We Built a Furniture Registry (and Why More Niches Need One)

Try shopping for a sofa online without already knowing the brand you want. The first page of Google is a mix of affiliate roundups ("Best 27 Sofas of 2026"), retailer ads, and category pages so vague that "modern" and "contemporary" mean the same thing on one site and opposite things on the next. Most "best of" lists are seeded by whoever paid for placement. Price signals are inconsistent — one store lists a 3-seater leather sofa for $1,800, the next for $4,200, and you can't tell whether you're paying for materials, brand, or just markup.

Search engines are good at the long tail. They're bad at giving you a curated, opinionated reference for a fragmented space.

Catalog exists because that gap doesn't fix itself.

What a registry does differently

A registry is not a search engine and not a marketplace. It is an editorially-maintained index. For us, that means a few specific things:

• Editorial selection. A store gets listed if it sells furniture as a primary product, has a real buying path, ships to at least the US/CA/EU, and isn't a dropshipping reseller. AI tools pre-fill suggestions; people approve.
• Fixed classification axes. Every store is sorted by style, price band, and a small set of attributes (trade programs, outlets, quick-ship). Price bands are benchmarked against a single reference product — a 3-seater leather sofa — instead of being self-reported.
• Liveness checks. Every URL is checked weekly. Stores get delisted within 14 days of a permanent failure.
• Transparency about money. Some outbound links pay us a commission. None of those payments influence whether a store is listed or how it's classified. The full rules are on the methodology page.

None of this is technically hard. It's just unrewarded by the search-engine-optimised content economy, which is exactly why nobody else does it.

This is a general pattern, not a furniture pattern

Any space with three properties tends to produce the same kind of broken search experience:

1. The underlying inventory is large enough that you can't memorise it.
2. The quality bar across that inventory is uneven.
3. The top results are dominated by SEO-optimised aggregators rather than reference.

Indie SaaS has the problem. Archival film has it. Audio gear has it. Anything that hobbyists know cold and newcomers find impenetrable. In each case, the answer is not "better search" — search is already working as designed. The answer is a small, opinionated index that someone is willing to maintain.

Sibling project: PeptideStack

The same problem exists in peptide research, where information is scattered across forum posts, vendor pages, and PDFs of varying provenance — and the search results are full of vendors with an obvious incentive to be the loudest voice on every compound.

That's why we also build PeptideStack — an independent index of peptides with structured profiles, maintained editorially rather than crowd-sourced. Different domain, same model: opinionated curation, consistent classification, no pay-to-rank. If you've ever tried to find a clean, comparable summary of a specific peptide and ended up reading six conflicting forum threads instead, that's the gap PeptideStack is for.

What makes the model work

Two things, mostly. The team is small, which means the scope stays narrow and the editorial bar stays consistent. And we're transparent about how we get paid, which means we can keep saying no to listings that don't belong here without anyone having to argue about it.

We're also willing to delist. That sounds obvious, but it's the part that pure SEO sites can't do — every page they remove costs them organic traffic, so they keep stale and bad listings live indefinitely. A registry that won't prune isn't a registry; it's just a longer search result.

If you run a furniture store you think should be on Catalog, or you spot a listing that's wrong, write to <a href="mailto:hello@catalog.homes">hello@catalog.homes</a>. We reply within a week and we log every correction.