Get in Touch
News
| Announcements

How Google’s AI search is changing
rankings, traffic and clicks

Fine white grid seen close up, representing the structure behind search results

Google now answers a growing share of searches before anyone clicks. An AI Overview sits above the results, summarises the answer and cites a handful of sources.

Rankings still matter. What has changed is what a ranking is worth, and which pages get chosen as the source.

Exposed mechanical watch movement, representing the detail beneath an AI answer

You may already see it in your own reporting. Impressions are steady or rising. Clicks are flat. Informational pages that once brought in steady traffic have gone quiet, even though their positions have barely moved.

It is tempting to read that as SEO getting weaker. A better reading is that search visibility is being redistributed towards businesses that give AI systems something specific and trustworthy to cite.

The businesses losing ground are usually the ones whose content says what everyone else says. The ones gaining ground have something specific worth quoting.

What changed on the results page

Search used to hand out ten links. It now hands out an answer, then the links that informed it.
Dark sculptural form against a pale background

For broad, informational questions, the AI Overview often satisfies the searcher on the page. A question like “what is equipment finance” may never produce a click, no matter who ranks first.

That traffic was always the least valuable. It was reading, not buying.

Traditional SEO compared with AI-driven SEO

For specific, commercial searches the picture is different. Someone looking for excavator finance for a two-year-old ABN still needs a specialist, and the pages cited in the answer are the ones they click.

  • Fewer clicks on general questions
  • More weight on pages that answer a precise need
  • Mentions and citations count, even without a click
  • The same content now feeds Google, ChatGPT and Gemini
The click has moved further down the funnel. That is bad news for thin content and good news for businesses with real expertise to publish.
The measurement gap
Why your reports can look worse while your position improves

Most SEO reporting was built for a world where a ranking led to a click and a click led to a lead. AI search breaks the middle of that chain, so the usual numbers mislead in three ways:

  • Traffic undercounts visibility. Being cited in an AI Overview, or recommended in ChatGPT, builds familiarity that never shows up as a session.
  • Average position hides the change. A top three ranking beneath an AI Overview can earn fewer clicks than a lower position did two years ago.
  • Branded search becomes the leading indicator. People who meet you in an AI answer often search your name later, and that is where the value shows up.

The better approach is to measure what AI search actually rewards: qualified enquiries, growth in branded search, the commercial terms you are visible for, and whether AI tools name you when asked about your category.

How this fits alongside paid media is covered in why your marketing isn’t working together.

What gets cited

What AI search rewards

AI systems choose sources the way a careful researcher would. They favour pages that are specific, consistent and clearly written by someone who does the work.
The pages that get chosen
01.
Answer one precise question better than anything else indexed
02.
Carry real detail, such as prices, criteria, timelines and examples
03.
Are backed by consistent business information across the web
04.
Sit within a site that covers its topic in depth
05.
Are clearly structured, with schema, headings and plain language
AI-driven SEO signals

Millard Financial, a Gold Coast equipment finance broker, started with fewer than 10 indexed pages while national lenders owned every broad search term.

Rather than compete for “business loans”, we built more than 20 pages on the specific equipment buyers actually search for, from excavator and crane finance to medical and aviation equipment. Within twelve months organic users grew from 12 to 669, the site ranked for more than 3,280 keywords, and Millard began to be recommended in ChatGPT and Gemini answers for its niches. Read the Millard Financial case study.

Specific beats general. AI search simply made it obvious.
Two approaches
Content written to rank vs content written to be cited

The same topic, written two ways. Only one of them holds its value in AI search.

Written to rank/Targets the broadest keyword available/Repeats what the top results already say/Written to a word count/General advice any business could publish/Measured on traffic
Written to be cited/Targets the precise question a buyer asks/Adds detail only you can provide/Written until the question is answered/Grounded in real cases and numbers/Measured on qualified enquiries
General content is exactly what an AI Overview replaces.
Specific content is what it needs to quote.
The response

Four ways to adapt your SEO for AI search

None of this replaces good SEO. It is good SEO, applied with more discipline.

1. Go narrow and deep

Map the specific commercial questions your buyers ask and give each one its own page. Millard’s growth came from twenty niche pages, not one page about finance.

2. Publish what only you know

Pricing ranges, eligibility criteria, timelines, common mistakes, real outcomes. AI systems cannot find this detail elsewhere, so it is the detail they cite.

3. Make your business unambiguous

Keep your name, address and service details consistent across your site, Google Business Profile and directories, and use structured data that states plainly who you are and what you do. AI systems quote sources they can verify.

4. Measure the right outcomes

Track qualified enquiries, branded search and AI mentions alongside rankings. Traffic alone will understate your progress. This is how we approach SEO for every client.

The goal is to become the obvious source in your niche, for people and for machines.
Exposure
How exposed is your current traffic?

A quick way to judge how much of your organic traffic AI search is likely to absorb.

More exposed if/Most traffic comes from general how-to content/Your pages read like the top five competitors/Leads come from broad, high-volume terms/Little of your content could only come from you/Few people search for your brand by name
Less exposed if/Traffic comes from specific commercial pages/Your content includes original data or cases/Enquiries come from niche, high-intent terms/Your expertise is visible on every page/Branded search is growing
If the left column sounds familiar, the traffic you are losing was probably low value. The priority is building the specific pages that replace it.
The opportunity
Polished chrome disc on a white background

Fewer clicks, better ones

AI search has not made SEO less valuable. It has made generic SEO less valuable, and removed much of the traffic that never converted anyway.

For a specialist business that is an advantage. Large competitors can outspend you on broad terms, but they rarely match you on the niche questions your best clients ask. Built properly, that content ranks in Google, gets cited in AI answers and strengthens every other channel you run. It is the marketing flywheel applied to search.

Want to be the source AI search cites in your niche?

Tell us where the business is heading and we will map what a connected marketing system looks like for it.
Start a project