The plain statement AI needs on your homepage

The plain statement is one sentence on your homepage that says what your product is, who it serves, how it is priced and what it replaces. Assistants need those four facts before they can make a specific recommendation.

Most homepages make the reader assemble the answer. They use a slogan, a broad promise and a button, then leave the useful facts on several other pages. In the 121 sites covered by the free AI readiness check, the complete statement check passed on 20 sites, failed on 43 and warned on 56.

The sentence buyers and assistants both need

The sentence is a compact buying label. It should identify the product, the intended customer, the price or pricing route, and the alternative it replaces. These are four separate facts, but they belong together because a buyer should not need to reconstruct the offer from a navigation menu, a pricing page and a comparison page.

This is different from a tagline. “Make work flow” may sound pleasant, but it does not tell a buyer what the product does. “A workflow tool for small agencies, with pricing shown on the plans page, instead of shared spreadsheets” gives an assistant usable material and gives a busy person a reason to read on.

The wording does not need to sound as if a machine wrote it. Use the terms your buyers use, keep the audience specific enough to be useful, and name a real alternative such as spreadsheets, an agency, an in-house process or a product category. The sentence should make the offer legible before the reader starts hunting for evidence.

That is the core of homepage clarity for AI. The assistant can still use your product page, pricing page and comparison pages, but the homepage should establish the basic facts first. Treat the sentence as a sign above the shop, not as a slogan printed on the window.

Where the gap appears

The plain statement was the weakest of the checks in this audit. The other weak checks were comparison or alternatives pages, and questions answered on the page. Here is the comparison table for those three checks:

The numbers point to a content problem rather than a simple access problem. AI crawlers were allowed in robots.txt on 116 sites, a clear H1 passed on 101, and the homepage answered a plain request on 119. A page can therefore be reachable and technically tidy while still leaving its position in the market unclear.

Product or SoftwareApplication schema passed on 60 sites and failed on 59. Organization schema passed on 72 and failed on 47. Nothing in the audit suggests structured data substitutes for the plain-language sentence. Schema and the statement check measure different things.

The point is not to replace technical work with copy. It is to stop treating technical work as proof that the offer is understandable. A crawler can reach a page that still makes a buyer do detective work, which leaves the buyer doing work the homepage should have done.

Build the statement from four facts

Start with the product category. Use a noun a buyer would recognise, such as accounting software, a research service or a project management tool. Avoid making the first words a benefit that could describe almost anything. Add the audience in terms your buyers use, such as “for small ecommerce teams”, rather than “for modern businesses”. If you serve several groups, choose the group the homepage is meant to convert and give the other groups their own pages.

State the price or the route to the price. If the price varies, say what controls it and link to the pricing page. The price-visible or pricing-page check passed on 90 sites and warned on 29, so this detail is worth checking even when there is no single price. “Talk to sales” on its own leaves too much unresolved for a buyer or an assistant.

Name the alternative. This can be a product category, a manual process or a familiar incumbent. Do not claim that every competitor is worse, and do not force a competitor's name into the sentence if buyers mainly switch from spreadsheets or an in-house process. Say what the buyer uses today and why the switch is worth considering, then use a comparison page for the supporting detail.

Use this structure:

[Product] is a [category] for [audience], priced at [price or pricing route], instead of [alternative].

The bracketed parts are prompts, not copy to publish. Replace them with facts and keep the sentence readable. If you cannot fill one part without vague language, that is a positioning problem as well as a homepage problem. The sentence should be short enough to scan, but complete enough to remove the first round of guesswork.

A worked example with the audit figures

Imagine, not one of the audited sites, a client reporting tool aimed at small agencies. The sentence might read: “A client reporting tool for small agencies, with pricing shown on the plans page, instead of building reports in spreadsheets.” Now walk through what the statement check means, line by line, without treating the invented product as an audited result.

Line 1, the 20 that passed. These sites met the complete test: the page made the product, audience, price or pricing route, and alternative clear in one plain statement or a close equivalent. Our hypothetical example would sit in this group if all four facts were stated clearly.

Line 2, the 43 that failed. These sites did not provide the complete statement. A homepage might name the product but omit the audience, hide the pricing route, or say nothing about what the buyer uses instead. The audit records the failed check, not which of the four facts each site missed.

Line 3, the 56 that warned. These sites gave some useful information, but the wording was incomplete or ambiguous enough to need a warning. A line such as “Reporting made simple for modern teams” gives a benefit, but leaves the category, audience, price and alternative unresolved.

That is the larger issue exposed by the audit. The test is not asking for a clever slogan or a long explanation. It is asking whether the page gives a buyer enough plain context to understand the offer without assembling it from other pages. The worked sentence does that in one place, while the warning example leaves the reader to guess.

Place it where the page can be read

Put the sentence in the first useful block of the homepage, close to the main heading and before a visitor has to open a menu. The heading can be shorter, but the supporting copy should carry the four facts. Keep the wording in normal page text, rather than placing the only version inside an image, animation or interactive component.

Repeat the core facts on the product page and pricing page, with more detail where it helps. Link the sentence to the relevant pricing route, then link to a comparison or alternatives page that explains the trade-offs. In this audit, comparison or alternatives pages passed on 54 sites, so this is a sensible place to add evidence.

Answer the obvious follow-up questions on the same page where possible. Who is it for? What does it replace? What does it cost? What does it not do? The questions-answered check passed on 56 sites and warned on 63, which makes a short, specific buyer section worth checking.

Keep the important words in the rendered page. The content-present-without-JavaScript check passed on 92 sites, failed on 18 and warned on 9. The point is not that every homepage must be plain text. It is that the key statement should remain available when the page is fetched without the visual layer.

Then check the supporting signals. Title and description saying what the product is passed on 92 sites, while canonical and Open Graph tags passed on the same number. A sitemap passed on 100. How CiteGraph measures explains these checks, but the checks do not repair unclear positioning.

Use the free AI visibility checker after the page has changed if you want to examine wider visibility work. Keep that result separate from the homepage statement check. A page can be clear and still not be cited, or cited for a reason unrelated to the sentence.

What this audit cannot establish

We can tell you how the audited sites performed on these checks. We cannot tell you that adding one sentence will make an assistant recommend your product, increase conversions or cite your homepage. We measured what fifteen fixed checks capture. A different audit, or a different set of assistants, could find something else.

The data also does not tell us which wording each site used, which assistants read each page, or whether a failed check caused a missed recommendation. It shows signals recorded in a fixed audit, not a controlled test of homepage copy. The pattern identifies a repeated weakness, not a guaranteed outcome from one edit.

That limit should make the work more practical, not less. Start with the sentence because it is cheap to inspect and easy for a buyer to understand. Then check the product page, pricing route, comparison content, crawler access and rendered text. Measure citations separately from readiness, because the audit does not establish that passing one causes the other.

The opening scene was a homepage that made the reader assemble the answer. The change is one clear label: what the product is, who it serves, how its price is handled and what it replaces. The assistant may still need more evidence, but the first question no longer has to be solved by guesswork.