AI Overviews Are Eating Your Informational Clicks. Here's What to Do About It.

What's actually happening to traffic

AI Overviews replaced Search Generative Experience in mid-2024 and have rolled out aggressively across Australian search results since. They now appear on a large share of informational queries - "what is", "how does", "why do", and most definitional searches.

The clicks haven't disappeared evenly. The pattern we see in client Search Console data is consistent:

• •Pure informational pages (definitions, "what is X" articles) - 30-60% CTR drops on the same ranking position
• •Comparison and "vs" content - down 10-30%, but the click that does come through is higher intent
• •Commercial and transactional pages ("hire X", "buy X", "X near me") - largely unchanged or slightly up
• •Local pack and Maps - unchanged

In other words: the AI is eating the easy clicks. The clicks that survive are the ones where the user still needs to evaluate, decide, or buy.

Step 1 - Audit which pages are bleeding

Open Search Console and compare the last 90 days against the same period a year ago. Filter to queries that contain "what", "how", "why", or "is". Look at clicks per query, not just impressions - impressions stay flat while clicks fall, which is the AI Overview fingerprint.

Rank the affected pages by traffic loss in absolute clicks, not percentage. A 40% drop on a small page matters less than a 15% drop on the page that used to drive a third of your organic.

Step 2 - Be the page the AI cites

AI Overviews cite sources. Being one of them keeps your brand visible even when the user doesn't click. The content patterns that get cited are predictable:

• •A direct, factual answer in the first paragraph - 40-60 words, no preamble
• •Clear H2/H3 structure that maps to the question being asked
• •Specific data points, numbers, and named sources rather than vague claims
• •A named author with credentials visible on the page (E-E-A-T)
• •Schema markup - Article, FAQPage, HowTo, Person
• •Content updated recently - AI Overviews lean heavily on fresh sources

Rewrite your top-bleeding pages with the answer in the first 40-60 words. The rest of the article can elaborate, but the AI is reading the top of the page hardest.

Step 3 - Shift your content mix

If informational queries are losing clicks, write fewer of them and more of what still converts. Three content types still pull clicks reliably:

Comparative and evaluative content

"X vs Y", "best X for Y", "is X worth it" - users still click these because they want to see the reasoning, not just the verdict. The AI gives a flat summary; the click goes to whoever shows their working.

Local and "near me" content

AI Overviews don't replace the local pack. Service-area pages, suburb-targeted content, and "best X in Perth" style queries still convert.

Bottom-funnel transactional content

Pricing pages, "hire a X", "get a quote for X" - these queries still resolve in clicks because the user needs to act, not just learn.

Step 4 - Treat brand mentions as a real metric

If the AI is answering the question, the win is being the brand mentioned in the answer. Track which queries surface your brand in AI Overviews and ChatGPT Search. The next 18 months will see this metric become as important as ranking position - and most agencies aren't measuring it yet.

Tools that help: Profound, Otterly, AthenaHQ, and direct testing. We track this manually for clients - the tooling is still maturing.

What we tell clients to stop doing

• •Stop writing thin "what is X" pages with no unique angle - the AI does that better and faster
• •Stop padding intros with throat-clearing - lead with the answer or get skipped
• •Stop ignoring schema markup - it's no longer optional for AI visibility
• •Stop measuring SEO success on clicks alone - track citations, mentions, and downstream conversions