AI Overviews and the New Click Economy

The answer moved to the top of the page

AI Overviews changed the deal at the top of the search results, and a lot of teams are still reporting on a click that Google quietly stopped sending. When someone asks a question now, an AI Overview often answers it inline, in a synthesized block above the old blue links, with a handful of sources named off to the side. The user gets what they needed without a visit. That is not a glitch to wait out. It is the new shape of the click economy, and the brands that adapt to it first will own the surface while everyone else argues about whether it is fair.

I have spent fifteen years moving numbers in large programs, including scaling revenue at a national lender and an estimated nine figures of incremental revenue for an automaker. The pattern is always the same when a channel rewrites its own physics: panic, denial, and then a small group of teams who simply re-engineer for the new rules and pull ahead. Let us be in that group.

What AI Overviews actually changed

It helps to be precise about what broke, because the fix follows from the diagnosis.

• The answer became the destination. For a wide and growing band of informational queries, the user's need is fully met inside the Overview. No click is required, considered, or earned. This is the zero-click reality at scale, and it is permanent for those query types.
• Citations replaced positions. Instead of fighting for rank six, you are now competing to be one of the three or four sources the Overview names. A citation inside the answer is worth more than a ranking on a page the user no longer scrolls to.
• The surviving clicks got better. This is the part teams miss while they mourn the volume. When the Overview handles the easy lookups, the clicks that still come through are from people who want depth, want to compare, or want to buy. Lower count, higher intent. That is a trade most businesses should take happily.

The strategic reframe: AI Overviews are not stealing your traffic so much as redistributing your expertise, with attribution, to people who would have bounced anyway. Your job is to be the source it pulls from.

How to earn presence inside the Overview

You cannot buy your way into an AI Overview, and you cannot trick your way in either. You earn it by being the cleanest, most quotable source on the question. That is an engineering problem with executable steps.

Answer the question in the first two sentences

Overviews are assembled from passages, not whole pages. A model extracts a chunk that already contains a complete, self-contained answer. So lead with the answer and support it afterward. "A standard home inspection takes two to three hours for an average single-family home" is extractable. "Inspection timing depends on a variety of factors we will explore below" is not. Invert the structure you were taught: conclusion first, evidence second.

Make every claim discrete and sourceable

• State facts as standalone sentences a model can lift without surrounding context.
• Attach a date and a basis to anything numeric or time-sensitive so it is safe to quote.
• Put your strongest, most original claim where a crawler will find it fast: the opening of the section it belongs to.

Structure for chunking

• Use H2s and H3s that read like the questions real people ask.
• Favor short paragraphs, definition lists, and tables. Tables in particular get pulled into Overviews cleanly.
• Give each section one job: one question, one well-formed answer.

This is the heart of the discipline I cover in my piece on generative engine optimization, and it applies to every AI surface, not just Google's.

Still earning qualified clicks when Google answers first

Presence inside the answer is the new ranking. But you still want the click when the click is worth having, and Overviews leave room for that when you set the table right.

• Win the queries Overviews cannot fully answer. Comparisons, decisions, configuration, pricing, and anything that depends on the reader's specific situation still drive clicks, because the model knows it should not guess. Concentrate effort there.
• Make the citation irresistible. Even inside an Overview, a named source that promises more depth, a calculator, or original data earns the curiosity click. Give the reader a concrete reason that the full page is worth the trip.
• Own the bottom of the funnel. Transactional and high-consideration queries are where the surviving clicks convert. This is why pairing AI visibility work with disciplined conversion rate optimization for organic traffic pays off twice: fewer visits, but each one matters more.

A framework: the PACE model for the Overview era

When I need a team to hold one idea in their heads, I give them a model. For AI Overviews, use PACE: Presence, Answer, Citation, Edge.

• Presence. Are you crawlable and structured well enough to be eligible for the Overview at all? If the machine cannot read you cleanly, none of the rest matters.
• Answer. Does a single passage on your page fully resolve the sub-question, in the first two sentences, with no hunting required?
• Citation. Are your entities, authorship, and organization unambiguous, so the system can attribute the claim to a named, trustworthy source?
• Edge. Does the page offer something the Overview cannot contain, original data, a tool, a decision aid, that converts the citation into a qualified visit?

Score each priority page one to five on all four. The lowest number is your next task. Simple beats clever, because the teams that ship outrun the teams that theorize.

Stop measuring this the old way

The hardest part is organizational, not technical. Your click graph can dip while your actual influence grows, and if sessions are the only number on the dashboard, you will defund the exact work that is winning. You need a new scoreboard, which I lay out in full in my post on measuring SEO when the clicks fall.

The short version:

• Track impressions and how often you surface inside Overviews, not just clicks.
• Treat citation share across answers like a ranking report, because it is the new ranking report.
• Watch branded search and direct traffic. When people see you cited and then come looking for you by name, that lift is your Overview visibility showing up in a number a CFO will respect.
• Move budget conversations toward assisted and multi-touch attribution, because the Overview that introduced you almost never gets the last click.

Your Monday-morning checklist

• Identify your ten highest-value queries and check which ones already trigger an Overview.
• Rewrite the opening of each target page so the core answer lives in the first two sentences.
• Add or repair schema so machines read your meaning without guessing, the way I describe in my work on schema markup as a translation layer.
• Define and standardize your key entity names sitewide so citations attribute cleanly.
• Reallocate effort toward decision, comparison, and transactional queries Overviews will not fully answer.
• Stand up a citation-and-impression report alongside your clicks report, today.

The bottom line

AI Overviews did not end SEO. They changed the unit of visibility from a ranked link to a cited claim, and that rewards the things great marketing always rewarded: clarity, structure, and genuine expertise a machine can trust. Numbers over noise, honest over hype. The brands that win the next few years will be the ones who became the most quotable source on the questions they care about, and made that easy for the machine to see.

I am writing one of these every week, working through what is actually moving the numbers in AI-era marketing, no hype. If this is the kind of problem you are wrestling with, the channel's open by introduction.