Key takeaways
- AI Max Search expands paid eligibility into longer, conversational queries that old keyword structures did not monetise.
- The main risk for UK advertisers is budget dilution, especially where tracking counts weak leads or low-margin sales as valuable.
- Advertisers should test AI Max in controlled campaign groups with separated budgets, strong negatives and qualified conversion imports.
- Google’s incrementality claim needs proving against profit, lead quality and cannibalisation, not campaign-level conversion volume alone.
- Cleaner landing pages, product feeds and offline conversion data now matter as much as keyword lists.
AI Max Search is not just another automation layer. It is Google turning longer, messier, conversational searches into paid inventory at scale. For UK advertisers, that means the next cost increase will not always come from higher CPCs on the keywords you already track. It will come from Google finding new ways to match your ads to queries you never chose. That is why our advice on AI-era Search ads keeps coming back to signal quality, tracking and control.
The commercial tension is simple. Google is creating more auctions. Advertisers are being told those auctions are incremental. Some will be. Plenty will be weak matches dressed up as reach. If your account already struggles to separate profitable demand from attractive noise, AI Max Search will magnify that problem.
The winners will not be the brands that switch everything on fastest. They will be the brands that know which queries, products, locations and leads deserve budget before Google starts widening the net.
What has changed in AI Max Search
Google says AI Max is now out of beta and has been adopted by more than 500,000 advertisers. The headline claim is bigger than feature adoption. Google says AI Max is opening billions of previously unmonetised searches to ads by matching against complex and ambiguous queries that traditional keyword targeting did not handle well.
The mechanism is Gemini-led intent interpretation. Instead of relying on a keyword list to define eligibility, Google is using AI to understand longer searches, infer commercial intent and match ads to a broader set of user needs. Google also says advertisers using AI Max or Performance Max see an average 15% lift in conversions or conversion value at a similar return on ad spend. Shopping relevance for complex queries is said to have improved by about 20%.
Alongside that, Google is testing ads inside AI Mode experiences, including Highlighted Answers, contextual sitelinks and Direct Offers. That matters because AI Max Search is not only about better query matching in familiar Search results. It is part of a wider move towards ads being inserted into AI-assisted planning journeys.
Why this matters for advertisers
The money moves because query eligibility expands before advertiser visibility catches up. In a normal keyword-led Search campaign, you at least know the core terms you intended to buy. With AI Max Search, Google is pushing further into inferred intent. The system sees a messy query, reads the landing page, assets and feed, then decides whether your ad is relevant enough to enter the auction.
That creates genuine opportunity for advertisers with clean data. If you sell a product or service where buyers search in detailed phrases, not neat keywords, AI-driven matching will pick up demand that old match types missed. Think legal services, B2B software, specialist ecommerce, travel, insurance, home improvement and local service categories. People do not always search in tidy noun phrases. They describe the problem. Google wants to monetise that description.
The risk is budget dilution. A broader matching system needs stronger boundaries. If your conversion tracking counts every form fill as equal, Smart Bidding will chase cheap signals. If your landing pages are vague, Google will infer vague intent. If your product feed contains thin titles, missing attributes and poor margins data, Shopping-style matching will steer spend towards easier clicks, not better profit.
Incrementality will become the argument
Google will position the new inventory as incremental. Advertisers should challenge that at query, audience and margin level. Incremental traffic means new profitable demand you were not already capturing through exact, phrase, Shopping, Performance Max or brand activity. It does not mean extra clicks that convert once, assist poorly, cannibalise brand, or bring in lower-quality leads.
This is where UK accounts are already vulnerable. Our Q2 UK spend report, based on PPC Geeks keyword-CPA analysis across 49 accounts in Q2 2026, shows the median UK account puts about 24% of search spend into keywords that convert nothing or cost more than three times its own average, up from 20% in Q1 2026. That figure rests on each account’s own conversion data, which is distorted in many accounts. Some of that spend is legitimate, including brand defence, tests and assists, so we frame it as spend not pulling its weight on last-touch, not pure waste. The rise is confirmed like-for-like: in the 42 accounts measurable in both quarters, median waste rose about 2 points, with 20 accounts getting worse and 13 improving. It is not a sample-mix artefact. If anything, 24% understates the issue because accounts with trustworthy tracking show a median of 36%.
That is the warning. If a standard keyword account already lets a quarter of spend drift into weak last-touch performance, a wider AI Max Search setup gives that drift more room. The answer is not to reject automation. The answer is to stop feeding it blunt signals.
PPC Geeks’ View
The specific problem advertisers will face is hidden query expansion inside accounts that already look healthy at campaign level. Performance will not fail neatly. It will blur. Search terms will look broadly plausible. Conversion volume will hold for a while. CPA will creep upwards. Sales teams will start saying lead quality feels softer. Ecommerce teams will see more revenue, but margin will lag.
We see this most often in lead-gen accounts running broad match with Smart Bidding and thin offline conversion feedback. Google is very good at finding more people who will fill in a form. It is less good at knowing which form fills became qualified opportunities unless you send that data back. AI Max Search increases the importance of that feedback loop because the system has more freedom to interpret intent before the click happens.
AI Max does not remove the need for account control. It moves control away from keyword lists and into tracking quality, landing page specificity, feed structure and budget rules.
— May Dayang, Digital Marketing Coordinator, PPC Geeks
For ecommerce, the practical observation is different. We see weaker outcomes when product feeds are built for catalogue completeness rather than commercial priority. If every product is eligible, Google will find demand for products that sell easily but return badly, run at poor margin, or distract budget from higher-value ranges. That is not an AI problem. It is an input problem.
This is exactly the type of issue we look for when you request a free PPC audit, especially where automation, tracking or campaign structure is affecting performance.
What advertisers should do next
Do not treat AI Max Search as a switch to flick across the whole account. Treat it as a controlled expansion test with strict entry rules, clean measurement and a clear definition of incremental value.
1. Split tests by commercial intent, not by convenience
Run AI Max against campaign groups where you can judge quality properly. For lead-gen, start with one service line where sales can mark leads as qualified, unqualified and closed. For ecommerce, start with a product category where margins, stock and returns are known. Do not test it on the messiest part of the account and call the result inconclusive.
2. Tighten tracking before widening matching
Check conversion actions in Google Ads before enabling wider matching. Remove soft actions from bidding, including page views, low-value form starts and duplicate calls. Import offline conversions for lead-gen accounts, with qualified lead and closed deal stages separated. If your sales cycle is longer than seven days, set up enhanced conversions for leads and offline uploads before asking Smart Bidding to find better demand.
3. Build negative rules from real economics
Create negative keyword lists based on poor-fit language, not just irrelevant terms. Add exclusions for jobs, free, templates, DIY, complaints, definitions and support searches where they do not match your commercial model. For B2B, exclude student, salary, training and vendor login intent. These negatives stop AI-led matching from spending on people researching around your category rather than buying from you.
4. Separate brand, non-brand and discovery budgets
Do not let AI Max expansion sit in the same budget pool as brand defence or proven non-brand demand. Put discovery-style testing into its own budget, with a hard cap and a defined success metric. For accounts affected by wider automation, our advice on the Smart Bidding budget update is relevant here: capped campaigns and target settings decide where Google is allowed to push harder.
5. Audit assets, feeds and landing pages for intent clarity
Google is reading more than keywords. Rewrite asset groups, page titles, product titles and landing page sections so each campaign has a clear job. A landing page that tries to cover every service in one place gives AI a weak signal. A page that clearly maps problem, audience, offer, location and proof gives matching systems a much better commercial brief.
6. Use source evidence, but test against your own economics
The Search Engine Land report is the clearest summary of Google’s claim that AI Max opens billions of previously unmonetised searches. Use that as a prompt to test, not as a reason to trust extra volume without proof.
When you compare performance, use Google’s own Google broad match guidance to understand how wider matching already works, then isolate what AI Max adds above your current broad match and Smart Bidding setup. For ecommerce teams, compare the test against the controls described in Google’s Performance Max campaign guidance, especially where asset groups, feeds and goals overlap with existing campaigns.
If you are also planning for AI Mode placements, use a separate reporting view. Our guide to Ads in AI Mode explains why conversational placements need different expectations from traditional Search. A click from an AI-assisted answer is not the same buyer moment as a click from a typed product query.
What this means for your campaigns
AI Max Search is a revenue expansion play from Google and a control test for advertisers. The extra inventory will suit accounts with strong tracking, clear landing pages, structured feeds and honest profit targets. It will punish accounts that optimise towards cheap conversions, blended ROAS or campaign-level averages.
The immediate action is to define what counts as incremental before the system spends. Set a test budget. Exclude weak intent. Separate brand. Import qualified outcomes. Check whether extra conversions become extra profit. If they do not, the new searches are not growth. They are just new places for budget to leak.
Google is making Search broader and more conversational. Your account needs to become more precise, not less.
If you’d like a second pair of eyes on how this affects your account, our team offers a free Google Ads audit, with no strings. For the detail behind this, see Introducing AI Max for Search campaigns.
Frequently asked questions
What is AI Max Search?
AI Max Search is Google’s AI-led expansion layer for Search campaigns. It uses broader intent interpretation, landing page signals, assets and related data to match ads to complex queries beyond a traditional keyword list.
Will AI Max Search increase Google Ads costs?
AI Max Search will increase eligible auctions. Costs rise when advertisers allow that extra eligibility to share budget with proven campaigns without strict conversion quality, negatives and budget caps.
Should UK advertisers switch AI Max on immediately?
No. UK advertisers should run a controlled test first. Pick one commercial area, separate the budget, fix conversion tracking and judge results by qualified leads, margin or profit.
How do I prove AI Max traffic is incremental?
Compare AI Max against clean control campaigns. Exclude brand, separate discovery budget, check search term quality, measure assisted value and match conversions back to qualified sales or profitable orders.
Which accounts face the biggest AI Max risk?
Lead-gen accounts using broad match and Smart Bidding with thin offline conversion data face the biggest risk. Google will find more form fills, but without sales feedback it will optimise towards volume rather than quality.
