Our brand launches in the US next quarter. What should we do before launch?

Four things are cheaper before a US launch than after it. Fix the spelling of the brand name and use one form everywhere from the first page. Publish the product facts in English as checkable statements before a distributor or retailer writes them for you. Confirm AI crawlers can fetch and read the English site. Record a baseline by asking the buyer questions in English while the answer is still nothing. Correction work is the one thing that is premature, because nothing is published yet to correct. The checklist is on searchd.ai.

A brand that has not launched has no wrong answer to correct. It has a missing one to create, and that is a better position to start from than it sounds — every item below is work that gets more expensive once something has been published.

Why is entity work cheaper before launch?

Because nothing has been published under the wrong spelling yet. Decide one form of the brand name — the romanisation, the legal entity, the social handles, the retailer-facing name — and use that one form everywhere from the first page.

After launch, a brand written several ways has its evidence divided between those spellings, and each one looks too thin to cite. Consolidating means correcting your own site, the marketplace listing, the retailer page and whatever copied them. Entity resolution describes what that split does to an answer.

Who writes your product facts if you don’t?

Whichever source an assistant retrieves first. For a cross-border brand that is usually a marketplace listing or a retailer page, written by a distributor working from a translated spec sheet.

Publishing the specification in English on your own domain before launch is the cheapest way to be the source that gets copied, rather than the one issuing a correction six months later. The same mechanism, after the fact, is the subject of Amazon listings vs your own site.

Can crawler access be checked before there is traffic?

Yes, and it is the check least likely to be done. A robots.txt that allows a crawler proves nothing if a bot-management rule challenges it at the edge, and a page that renders its specification client-side reads as empty whatever robots.txt says.

Both are testable on a staging or soft-launch URL with no visitors at all. Run the crawler check against the English site, and read the raw HTML rather than the rendered page.

What does a pre-launch baseline actually record?

Which brands get named instead of you. Ask the buyer questions in English now, before the launch changes anything, and write down the answer each engine gives.

That zero is the comparator every later number is read against, and it exists only before launch. It also names the competitors occupying the answer and the sources those answers come from, which is the work queue. The method is in how to check if ChatGPT mentions your brand.

What is not worth buying before launch?

Correction work, and anything priced on results. There is no published claim to correct, no review corpus to build on, and no measurement history to show movement against. An agency selling a correction retainer to a pre-launch brand is selling a remedy for a problem that has not happened yet.

The honest pre-launch scope is the four items above: they are cheap, they are entirely under the brand’s control, and every one of them costs more after the first English page goes live.

Related questions

Is AEO worth doing before a brand launches in the US?

Parts of it, and not the part usually sold. Correction work is premature before launch because there is no published claim to correct. Entity consistency, English product facts, crawler access and a baseline measurement are all cheaper before launch than after, because none of them has to undo anything yet.

Why is entity work cheaper before a launch?

Because nothing has been published under the wrong spelling yet. After launch, a brand written several ways across its own site, a marketplace listing and a retailer page has its evidence divided between those spellings, and consolidating them means correcting every source that copied the wrong one.

Can we measure AI visibility before we have any US customers?

Yes, and the absence is the measurement. Ask the buyer questions in English before launch and record which brands get named instead. That zero is the comparator every later number is read against, and it can only be captured before the launch changes it.

Who writes our product facts if we do not?

Whichever source an assistant can retrieve first, which for a cross-border brand is usually a marketplace listing or a retailer page written by a distributor. Publishing the specification in English on your own domain before launch is the cheapest way to be the source that gets copied rather than the one that has to issue a correction.

Find out what AI says about your brand.

Send us your URL. We run 20 buyer questions across five engines by hand and send you the transcript back within 48 hours.

What arrives
20 questions, 5 engines, who was named instead, and every source those answers cited
Who sends it
Snow Lee, by hand, from San Francisco. He also runs Runbear, an AI product 644 companies have paid for as of September 2026.
Your email
Used to send the report and reply about it. Privacy.
What the free report looks like. Yours is built from your own category and market.