I rewrote one product page to directly answer its own H1. It started getting cited by ChatGPT within 2 weeks.

I picked one underperforming product page as a deliberate test case — not a redesign, just a rewrite of the headline and opening sentence to directly answer the exact question I knew people were searching. Nothing else on the page changed materially. The difference in AI-search visibility was bigger than I expected from such a small edit.

The Starting Point

The page's original H1 was branded and vague — the kind of heading that reads fine to a human skimming a site, but gives an AI system nothing concrete to match against a real question. The actual useful information — which configurations a specific product was compatible with — was accurate and present on the page, just buried a few paragraphs down under generic framing.

The Change

I rewrote the H1 to state the actual question as plainly as a searcher would type it, and rewrote the first sentence below it to answer that question directly, in the first few words:

Before:

<h1>Premium Replacement Components</h1>
<p>We offer a wide range of high-quality parts...</p>

After:

<h1>Replacement Part Compatible With Model X (2014–2019)</h1>
<p>This part fits Model X units from 2014 to 2019 in the standard
configuration. It does not fit the extended configuration — see
compatibility table below for exact matches.</p>

Everything else on the page — images, pricing, the rest of the description — stayed the same.

What I Tracked

I didn't have a formal before/after measurement tool running, so this is closer to a direct observation than a controlled experiment: I periodically asked ChatGPT and Perplexity variations of the actual question, and watched whether this page showed up as a cited source, checking every few days for about a month.

What Happened

Before the rewrite, the page never appeared in any of these checks. Around two weeks after the change, it started showing up — first in Perplexity, then ChatGPT shortly after. This is one page, observed informally, not a controlled study — I can't rule out other factors (general crawl timing, unrelated site changes) contributing, and I wouldn't present this as a guaranteed timeline for anyone else's page.

What I Learned

The lesson that held up on reflection: the actual factual content barely changed, but where the direct answer sat on the page, and how closely the heading matched a real search question, seemed to matter more than I expected going in. It's a cheap, fast thing to test on your own underperforming pages — rewrite the H1 and opening line to directly answer the specific question, leave everything else alone, and see what happens over a few weeks. It won't tell you why it worked with certainty, but it's low-effort enough to be worth trying before assuming a page needs a bigger overhaul.