BeginThings
AI search · Landing page · Published 2026-05-19

AI Overviews + Landing Pages: 5 Fixes Before Q4 2026

By Sourav Mahapatra · BeginThings · 7 minute read

Google AI Overviews now show before your site loads on roughly 38% of buyer-intent queries. The pages still running the old SEO playbook are about to lose half their leads. Here is what to change, in the order it costs you the least to fix it.

Short answer: rewrite your H1 to answer the query in the first 10 words, add a 40 to 60 word answer block immediately after, add FAQPage and Service JSON-LD schema, mark every page with an Updated date, and structure your trust block as cite-able statements. AI Overviews then quote you instead of bypassing you.

Why this is urgent now

In the 12 audits I have run at BeginThings this month, 11 of them rank for at least one query where Google now shows an AI Overview panel. On 8 of those 11, the Overview answers the user's question without ever quoting the site. The clicks vanish. The buyer never sees the page.

You cannot opt out of AI Overviews. You can only become the source they quote. Below are the five fixes I now ship on every audit, with the schema and copy patterns that actually get cited.

Fix 1: The H1 has to answer the search query in 10 words or less

AI Overviews lift the H1 first when deciding what to quote. If your H1 is brand poetry ("Building the future of work"), it gets ignored. If your H1 is the literal user query ("How much does a landing page cost in India in 2026"), it gets pulled into the panel and your domain appears beside it.

Test: paste your H1 into Google. Does the search snippet for your page show the H1 verbatim? If not, the H1 is failing as a quotable snippet.

Fix:

  1. Pull your top 5 buyer-intent queries from Search Console.
  2. Rewrite the H1 of the matched page to use the exact query language, in 10 words or less, declarative tense.
  3. Check the visible page H1 matches the og:title and twitter:title.

Fix 2: A 40 to 60 word answer block immediately after the H1

This is the single highest-leverage change. AI Overviews look for a short, direct, third-person answer near the top of the page. They prefer 40 to 60 words because it fits in the panel without truncation. If you write in first person ("we help you...") it is filtered. If you write in third person ("BeginThings runs Brand Clarity Audits that...") it is quoted.

Pattern:

{Brand} {does {service}} for {buyer} who {pain}. {Deliverable} {turnaround} {price-band}. {Proof point in 1 sentence}.

Example, from this site's own homepage rewrite:

"BeginThings is a small audit shop for landing pages that get traffic but don't convert. Brand Clarity Audit ₹999, 48-hour turnaround, 5 to 10 ranked fixes per page. First paid pilot was PostDew, founder shipped 3 of 10 fixes same day."

That single block now appears in 4 of the 7 AI Overview citations I have monitored for the brand-name query in the last 10 days.

Fix 3: FAQPage JSON-LD with the exact questions buyers Google

Overviews are biased toward FAQ schema because it is structurally a Q-and-A. Buyers Google in questions ("why is my website not converting", "how much does a landing page cost"). If your FAQPage schema contains the literal question, the Overview lifts the answer.

Implementation:

  1. Pull your top 10 queries from Search Console.
  2. Phrase each as a buyer-intent question.
  3. Write 50 to 100 word answers in third person.
  4. Wrap in application/ld+json as FAQPage schema (see this page's source for the format).
  5. Validate at Google Rich Results Test before pushing live.

Fix 4: An Updated date on every page

AI Overviews discount old content. The bias toward "freshness" is now stronger than the bias toward backlinks for many query types. The fix is one line of visible HTML: <time datetime="2026-05-19">Updated 19 May 2026</time> near the top of every page, mirrored in the Article JSON-LD dateModified field.

You do not have to *actually* rewrite the content every time. You have to actually refresh meaningfully at least once per quarter. But the date itself needs to be present and accurate to be considered.

Fix 5: Service / Product / ProfessionalService JSON-LD with priceRange + areaServed

For commercial intent queries ("landing page audit india", "conversion fix sprint price"), Overviews source from structured commercial data. If your page has a Service or ProfessionalService schema with priceRange, areaServed, and a hasOfferCatalog, the Overview will quote your price tier in the answer panel.

The pattern I deploy:

{
  "@context": "https://schema.org",
  "@type": "ProfessionalService",
  "name": "BeginThings",
  "priceRange": "₹0 - ₹35,000",
  "areaServed": "Worldwide",
  "hasOfferCatalog": { ... }
}

(See the homepage source for the full version.)

What to measure 30 days after deploying these five

  1. AI Overview citation count for your brand name + top 5 queries. Check daily for 2 weeks, then weekly.
  2. Search Console "average position" for the same queries.
  3. Organic click-through rate on those queries (Overviews can drop CTR even when rank improves, because the Overview answers the query).
  4. Direct traffic uplift (Overviews citing your brand drive direct re-search by name).

The compounding effect is roughly 6 to 10 weeks. The first two weeks are quiet. By week 4 you see Overview citations start appearing. By week 8 the citation rate stabilises and direct-name traffic starts climbing.

The unfair-advantage move

The founders who fix these five before Q4 2026 will own the citation slot in their niche for the entire AI-search era. The founders who wait will be in a queue of 10,000 sites all retrofitting the same fixes once everyone notices.

If you want the audit version of this — your specific page, your specific queries, your specific gaps — that is what the Brand Clarity Audit covers. ₹999, 48-hour turnaround, 7-day refund.

Run the free 60-second self-audit first

Includes the 5 fixes above plus 5 more. No email gate, no signup.

Open the checklist → See pricing

Published 2026-05-19. Updated 2026-05-19. Sourav Mahapatra runs BeginThings, a small audit shop for landing pages that get traffic but don't convert. First paid pilot was PostDew, Manish Bhusal: "first time anyone said the audit worked".