There are two search engines

Sooner or later every owner asks some version of the same question. How do I get my business to show up on DuckDuckGo? What about Yahoo? Should I be doing something for ChatGPT? The question sounds like it should have a long answer, one engine at a time. It doesn't. Almost none of those engines look at your website at all.

There are two working maps of the web that matter: Google's index and Microsoft's, the one behind Bing. Building one is so expensive that nearly everyone else rents. Yahoo runs on Bing's index. DuckDuckGo's web results come from Bing. Ecosia, Bing. Startpage, Google. When ChatGPT searches the web for an answer, it searches Bing. Copilot, Bing. The names on the doors differ; the building behind them is one of the same two.

Two gates, one contract

This is better news than it sounds. It means you do not have a DuckDuckGo problem or a Yahoo problem or a ChatGPT problem. You have, at most, a Google problem and a Bing problem, and both gates read the same paperwork: the same robots file, the same noindex instructions, the same sitemap, the same canonical tags, the same words in the page's first response. Pass both gates and you are present everywhere that rents from them, automatically, with no extra work. Fail one and a whole family of engines goes dark together.

That last part deserves a plain sentence, because it cuts the other way too. A single blocking line aimed at Bing's crawler, one line, in one file, added on purpose years ago or by a platform default nobody reviewed, removes you from Bing, Yahoo, DuckDuckGo, Ecosia, and the search behind ChatGPT at once. Nobody gets an email about it. The engines that rent an index cannot be blocked separately, which means they cannot warn you separately either.

The two that walk the site themselves

Two engines of note keep their own maps. Brave Search builds its index in an unusual way: pages become candidates when enough Brave users have actually visited them. So if your site is missing from Brave, that is usually a statement about foot traffic, not about your website. It is the one absence on this list that is not a defect, and anyone who diagnoses it as one is reading the wrong instrument.

Perplexity crawls for itself and works differently in a way worth understanding: it does not rank pages, it quotes them. When someone asks it a question, it pulls pages from its index, lifts the useful passages, and writes an answer with sources attached. There is no position two in an answer. Your page either gets quoted or it does not exist in that conversation.

What the AI engines actually read

Here is the part that changes how you should think about your own site. The crawlers behind the answer engines read the raw page, the first response your server sends, before any JavaScript runs. Google will wait for your scripts. These readers mostly will not. A site that loads as an empty shell and paints its words in later is not handicapped in an AI answer, it is absent from it.

I have written before about sites that look finished to a person and blank to a crawler. Every word of that gets more true each year, because each new reader that matters is less patient than the one before it.

The earlier article: your website looks finished. Google may see a blank page.

What can and cannot be measured

One honest limit to close on. The engines that rent an index also tend to be the private ones, and private engines publish no dashboards for site owners. There is no DuckDuckGo console, no Startpage report, no Brave analytics. The only trace they leave is a referrer line in your own analytics, and the privacy-minded ones trim even that. So visibility on those engines can be verified at the gate, from the outside: allowed in or blocked. How much traffic each one sends you is partly unknowable, by anyone, with any tool. That is not a gap in your setup. That is the product working as designed, and anyone who quotes you a precise DuckDuckGo number is guessing.

Two gates, one contract, a handful of independents, and a hard limit on what anyone can measure. That is the whole landscape. Checking your own standing at both gates takes about a minute, and the check below reads your site exactly the way the crawlers do: one request, no JavaScript, no login, no access.

Run the ten-check audit: see your site the way the crawlers see it