Is Your Business AI-Ready? A 10-Point Check

Your content exists in the HTML

Test: open your page, view source, and search for a sentence you can see on screen. If it is not in the source, your text is being assembled in the browser and some crawlers will see an empty page.

Fix: server-rendered pages. This is a build decision, not a plugin. Every site we ship renders its content on the server for exactly this reason.

Every service has its own page

Test: count your services, then count your service pages. If nine services share one page, you have nothing specific for an engine to cite about any of them.

Fix: one page per service you actually want calls for. At $50 per page past the first five, this is the cheapest visibility you can buy — and the page planner will tell you how many you need before you spend anything.

Your cities are named in text

Test: search your own site for the towns you serve. A map embed does not count — machines read words.

Fix: name your service areas in body copy, and give the towns that actually produce revenue their own page. “We serve the surrounding area” is invisible.

Prices or ranges appear somewhere

Test: can a stranger learn roughly what you charge without calling? If not, neither can a model, and cost is one of the most common things people ask AI about local services.

Fix: publish defensible ranges with the variables that move them. “$250 to $600 depending on door size and hardware” is quotable and it filters out bad-fit callers.

Headings are questions and answers come first

Test: read your H2s. Do they sound like the way customers ask, or like brochure labels — “Solutions,” “Our Approach,” “Why Choose Us”?

Fix: rewrite them as questions and answer each one in the first two sentences underneath. Engines lift openings; conclusions buried at the bottom of a section do not get quoted.

Structured data describes your business

Test: run a page through a rich-results validator. You want LocalBusiness, Service and FAQPage markup at minimum, with hours, location, phone and service list populated.

Fix: add schema during the build, and only mark up facts the page actually states. Markup is a label, not a substitute for content.

Your details match everywhere

Test: compare name, address, phone, hours and service areas across your site, your Google Business Profile and any directory you appear in. Look for old numbers and stale hours especially.

Fix: pick one canonical version and correct every source. Answer engines cross-check, and contradictions get you dropped rather than corrected.

Reviews exist and are recent

Test: how many reviews do you have, and when was the most recent one? Reviews are the main third-party evidence a model has that you are real and currently operating.

Fix: ask every finished job, in person, once. No incentives, no gating. Recency matters more than volume past a certain point.

The site is fast and technically discoverable

Test: check your Core Web Vitals on mobile, then confirm you have a sitemap.xml, a robots.txt that does not block crawling, and correct canonical tags. An llms.txt file is a cheap addition that lays out your site for AI crawlers.

Fix: most of this is a build-quality issue. We ship sites at an A score for SEO, AEO, UX and page speed, and the free audit will show you where yours actually stands.

How to score yourself honestly

Give yourself one point per check, and only count a pass if you verified it rather than assumed it. Most owners assume they pass items two, three and seven, and most fail at least one of them because a service was added or a phone number changed and the site never caught up.

Eight to ten means your foundation is sound and your remaining work is content and reviews. Five to seven usually means structure: missing service pages, missing city coverage, or copy with no specifics. Four or fewer almost always means the site itself needs rebuilding rather than patching, because the readability problems are baked into how it was made.

Whatever the score, write it down with the date. Six months from now that note is the only evidence you will have of whether anything improved.

The three fixes that move the score most

If you only do three things this quarter, do these — in this order.

• —Make the content readable without JavaScript. Nothing else on the list matters if a crawler sees an empty page.
• —Give your top three revenue services their own pages, each with a price range, a timeline and the towns you cover.
• —Rewrite the openings. Move the answer to the first two sentences under every heading and replace adjectives with numbers.

Those three cover checks one through five and most of six, and they are the same changes that improve conversion rate for human visitors. Schema, listings consistency and reviews are worth doing next, but they amplify a readable site rather than rescuing an unreadable one.

What this costs to fix

If the answer is a rebuild, ours is $300 for up to five pages, $50 per additional page, and $50 a month for hosting — with the technical foundation, schema, sitemap, llms.txt, answer-first copy and conversion tracking included rather than sold as SEO or AEO add-ons.

If the answer is a few more pages on a site that already renders properly, it may be a couple hundred dollars of work. The point of the checklist is to tell those two situations apart before anyone quotes you a retainer for either.

You can tell whether any of it works

Test: do you know how many calls and form submissions came from the site last month? Not sessions — conversations.

Fix: analytics plus conversion tracking on calls and forms, set up during the build. Then ask ChatGPT, Perplexity and Google your own service-plus-city questions once a month and write down who gets named. That log is your AEO scoreboard until better tooling exists.

Score yourself. Eight or more and you are in good shape. Five or fewer and the problem is structural, which is fixable in one build rather than in twelve monthly reports.