R1 · Reachable, can search engines and AI bots fetch, render, and index your site?

Turn JavaScript off in your browser and load your best product page. Whatever is still on the screen is roughly what an AI crawler gets. Whatever vanished, the spec table, the price, the comparison grid, is content no assistant will ever quote you on.

R1 · Reachable is the first of the five questions the Quattr Method asks, and the bluntest. Can search engines and AI bots fetch, render, and index your site at all? The other four stand on this one, because these are physical constraints rather than preferences. You can't be cited if you were never crawled.

Teams under a new AI mandate reach for citation dashboards first, the ones counting how often an AI answer names one of your pages as a source. A citation problem is often a fetch problem wearing better clothes, and this check starts further down.

No AI crawler executes JavaScript

Classic Googlebot renders pages. The crawlers feeding answer engines do not, meaning ChatGPT, Perplexity, Gemini and the rest of the tools that answer a question directly and cite their sources.

Picture a product page that builds its spec table in the browser. Google sees the table. GPTBot sees the empty container around it. That page can sit in Google's top results and contribute nothing to an assistant's answer. The crawler that came for it left with a shell.

So the method scores Reachable twice. Once for search engines, once for AI bots. The question is no longer whether your site is crawlable. It is whether the specific bots you need saw the specific content you want used.

Blocking one bot removes one surface

Bot access stopped being a single decision. GPTBot, ChatGPT-User, PerplexityBot, ClaudeBot and Google-Extended each open or close their own door. A robots.txt line written during an old scraping panic may still be deciding, quietly, which engines are allowed to consider you.

The bots also differ in what they are for. Some crawl to build training corpora. Others fetch a live page to answer a question in the moment. So allowing or blocking them is several separate decisions, each with its own cost on its own surface.

Which is why this check starts from what your server recorded. Which AI bots came, what they asked for, and what they got back.

Server logs are the ground truth

Quattr reads them joined to a crawl of your site, so bot behavior and page rules sit in one view.

One check reports per-bot fetch volume, status codes, and how many unique pages each bot reached. A second verifies the IP behind each crawler, so a spoofed hit never counts as coverage. Two more name the pages blocked from indexing and the pages nothing links to.

Crawl budget and page experience belong here too. One check finds the parameterized URLs eating fetches your money pages needed, and another reads the vitals that decide whether a fetch was worth the trip.

The server logs source page shows the data joined live, and the ai-crawler-readiness skill runs the whole check as one conversation: verified bot fetches, crossed against the pages you want cited. A skill is a packaged analysis you run by name in your assistant.

Freshness is part of the same check. Log data lands hours to nightly, vitals refresh on a slower cycle, and every answer carries its as-of date, the day that data was last complete. A score without one is a guess with confidence.

One kind of work owns this gate

On the method's cross-map, technical and crawl health is the primary lever on Reachable. Internal linking and site architecture assist, because a page nothing links to is unreachable in the plainest sense.

Sequencing is fastest-proof-first: run the work you can prove in weeks before the work that takes quarters. So the baseline sweep runs before the two fastest levers, internal linking and content refresh. Shipping improvements to pages the bots cannot fetch is effort spent below the gate.

Teams with a mandate from the top resist starting here. The mandate wants a citation answer and wants it now. The honest reply is that this is the fast part. It reads data you already own, and it either clears the ground above or finds the blocker that would otherwise have capped every later piece of work you funded, quietly, for as long as it sat there.

Diagnosis descends, plans start here

In practice you rarely start at the bottom. You notice citations sagging at R3 · Referenced, the question of whether you are the source answers use. You rule out ordinary week-to-week wobble. Then you work down the ladder until the explanation surfaces. Often it surfaces right here.

A blocked bot. A rendering gap. A page nothing links to.

The Quattr Method page maps all five questions and the eight kinds of work that move them. This is where a diagnosis can end, and where every plan begins.

See also: all five questions and the work that moves them →