The short answer: simplify the account, not the thinking

If you run a small business, you probably do not need a separate campaign for every close keyword variation anymore. AI Max for Search campaigns is designed to expand beyond the keywords you specify, using broad matching, keywordless targeting, ad text customization and final URL expansion. Google describes it as an optimization layer inside an existing Search campaign, not a new campaign type and not a replacement for Search itself. (support.google.com)

That makes simplification sensible. A smaller account can usually work with fewer campaigns organized around meaningful differences: brand versus non-brand, service or product line, geography, budget, and the quality or value of the conversion. Splitting campaigns because one keyword says “emergency” and another says “urgent” is becoming harder to justify.

The mistake is treating that as permission to put every service, audience and commercial objective into one automated bucket. AI Max can choose queries, copy and landing pages. It cannot decide whether a signed contract is worth more than a contact-form submission unless you give Google that signal. Your structure still needs to express the business decisions that automation cannot safely infer. When Attribution Is Unreliable, Price Marketing Around Qualified Opportunities makes the same point from the measurement side: optimize toward the opportunity that matters, not the easiest event to count.

Keyword control still wins when the cost of a bad lead is high

What is actually changing in September 2026

The phrase “AI Max is replacing keyword-by-keyword management” is directionally right, but technically loose. AI Max is not replacing every keyword campaign. It is absorbing several older expansion features into one system. Google says campaigns using automatically created assets or the campaign-level broad match setting are being automatically upgraded during September 2026. Dynamic Search Ads have a later transition, beginning in February 2027 after Google extended the timetable. (blog.google)

The migration matters because it changes where control lives. Instead of managing a collection of separate features, advertisers will increasingly manage AI Max settings such as search term matching, text customization, final URL expansion, brand controls and location-of-interest controls. Google says legacy settings will be carried into the new setup, with the aim of preserving performance and URL controls during the transition. (blog.google)

That is a reason to review your account now, before an automatic upgrade exposes weak landing pages, loose conversion actions or forgotten exclusions. It is not a reason to rebuild every campaign from scratch. If your current setup is profitable and your conversion tracking is trustworthy, preserve the parts that represent real commercial differences. Simplify the rest.

A safer way to simplify an account

1. Separate the economics first

   Keep campaigns apart when budgets, margins, locations, sales cycles or acceptable lead costs differ.

2. Keep a clean brand campaign

   Brand demand has different intent and often different economics from non-brand discovery. Do not let expansion blur the two.

3. Fix the conversion action before testing

   Use qualified calls, booked appointments, completed purchases or downstream CRM stages when those are the outcomes the business actually values.

4. Run AI Max as an experiment

   Google provides an AI Max experiment that divides traffic inside the existing campaign between a control and treatment rather than requiring a copied campaign.

5. Review search terms by source

   Separate existing keyword traffic from AI Max broad and keywordless matches so apparent growth is not mistaken for incremental growth.

6. Check generated copy and landing pages

   Approve useful assets, remove claims that do not fit the business and restrict URL expansion when important traffic must land on a specific page.

7. Scale only after quality holds

   Judge qualified opportunities, revenue or completed jobs over a meaningful period, not just clicks or form volume.

The numbers support testing, not blind adoption

Google’s launch data is attractive. The company reported that advertisers activating AI Max typically saw more conversions or conversion value at a similar CPA or ROAS, with a larger typical uplift among campaigns still relying mostly on exact and phrase match keywords. Google also published individual examples, including a utility connection company that reported more leads at a lower cost per action after enabling AI Max. Those are Google-selected results, so treat them as a reason to test, not as a forecast for your account. (blog.google)

Independent practitioner evidence is more mixed. Search Engine Land reported on tests across mature advertisers and found that the full feature set appeared more successful than search-term matching alone. The same analysis found a measurable improvement in weighted Quality Score after text customization was enabled, while warning that it remained too early to declare one ideal account structure. (searchengineland.com)

Other practitioners report a more familiar pattern: AI Max finds some useful incremental demand, but it also takes credit for conversions that might otherwise have arrived through existing keywords. One agency found AI Max using a modest share of Search spend while producing stronger ROAS than its non-AI Max traffic, yet also found overlap with existing short- and mid-tail queries. Its conclusion was practical: strong conversion signals improved results, while search-term audits and negative keywords remained necessary. (workingplanet.com)

That contradiction is not really a contradiction. AI Max may be good at finding cheaper or less obvious paths to a conversion and still be poor at distinguishing a valuable lead from a merely countable one. The outcome depends on the account’s inputs, the commercial value of the conversion and whether you measure incrementality at the campaign or account level.

The figures worth putting beside the decision

Small businesses have less room for bad automation

A national ecommerce advertiser can tolerate some wasted queries if the system finds enough profitable purchases. A local electrician, private clinic or specialist consultant often cannot. One irrelevant click is not just wasted media spend. It can consume staff time, trigger a poor customer experience and make an apparently healthy lead count look better than the sales pipeline really is.

That is why lead-generation businesses should be more cautious than ecommerce businesses. Search Engine Land’s testing notes that automated targeting can produce unqualified inquiries, especially when form submissions are treated as the main conversion. The publication’s practitioner tested AI Max across multiple accounts and specifically warned about lead quality and spam. (searchengineland.com)

The practical response is not to reject AI Max forever. It is to feed it better evidence. Import qualified opportunities back into Google Ads. Assign values to different lead stages where the sales process supports it. Exclude services, jobs, locations and customer types you do not want. Review competitor traffic and search-partner performance. Google’s own help documentation says AI Max is less effective when campaigns are budget-constrained, because expanded targeting cannot do much if the campaign cannot enter enough auctions. (support.google.com)

If your tracking stops at “form submitted,” simplify the campaign only after fixing that measurement gap. Otherwise, you are asking automation to optimize the wrong business.

The right operating model is fewer campaigns and better guardrails

Small businesses should simplify Google Ads now, but the simplification should happen at the level of business intent. Use fewer campaigns where the old structure only separated close keyword variants. Keep distinct campaigns where the economics or customer promise changes. Then test AI Max in the campaigns with reliable conversion data, enough budget headroom and landing pages that clearly explain the offer.

Do not measure success by how little time you spend inside the account. Measure whether the account produces better customers with less unnecessary management. AI Max can reduce the value of hand-building endless keyword lists. It does not reduce the value of positioning, useful landing pages, qualified conversion data or a human who notices that the “conversion” is the wrong customer.

The best small-business account in this transition will not be the one with the most keywords or the fewest. It will be the one where campaign boundaries reflect real decisions, automation is allowed to discover demand, and the owner still knows exactly what kind of customer the budget is meant to buy.