Maqtrix resources · 13 min read

Google AI Max Migration Checklist for Service Businesses: What Changes in September 2026

Audit Google AI Max before or after migration: location intent, brand controls, URLs, generated text, conversions, budgets and 30-day monitoring.

DIRECT ANSWER

Starting in September 2026, Google said eligible Search campaigns using automatically created assets or the campaign-level broad match setting would begin moving those features into AI Max. Dynamic Search Ads follow a separate timeline: Google revised their sunset and automatic transition to February 2027. Before evaluating performance, confirm which setting changed and audit location intent, brands, destination URLs, generated text, conversion goals, budgets and lead quality.

What changed in September 2026?

AI Max is an optimization layer inside Search campaigns. It is not a separate campaign type. Its two main groups of features are search-term matching and asset optimization.

Google's June 11 update to the migration announcement separates the current timelines:

Existing setupCurrent transitionWhat to inspect
Automatically created assetsEligible campaigns began automatic upgrades in September 2026Search-term matching and text customization settings
Campaign-level broad matchEligible campaigns began automatic upgrades in September 2026Search-term matching and query expansion
Dynamic Search AdsSunset and automatic transition begin in February 2027Dynamic ad groups, URL rules, generated responsive search ads and historical settings
New Search campaignGoogle Ads Help says AI Max is selected by defaultEvery feature toggle before publishing the campaign

The account notice and current interface remain the best evidence of whether a particular campaign is eligible or has already changed. Rollout timing can vary.

Google reports that non-retail advertisers using the full AI Max feature suite averaged 7% more conversions or conversion value at a similar CPA or ROAS than advertisers using search-term matching alone. This is Google internal 2026 aggregate data, not an independent study or a forecast for an individual account. More reach is useful only when the additional enquiries are relevant, serviceable and commercially sound.

WHAT GOOGLE ANNOUNCED TODAY — SEPTEMBER 23, 2026

Google announced that the closed beta of AI Brief is expanding to Dutch, French, German, Italian, Japanese, Portuguese and Spanish. Google also announced a unified reporting view intended to connect the search term, creative and landing page in one journey. Google said it will share more availability details later in 2026. These announcements do not mean either feature is available in every account today. Read Google's announcement.

Identify your exposure before changing anything

Create a one-page inventory of every Search campaign. Record the following:

  • Campaign name and business objective
  • Service and service area
  • Bid strategy and budget status
  • Text customization status
  • Campaign-level broad match status
  • Dynamic ad group usage
  • URL expansion and exclusions
  • Brand inclusions and exclusions
  • Primary conversion actions
  • Owner and last material change

Then classify each campaign:

  1. Already moved: AI Max is active or the account records an upgrade.
  2. September-eligible: automatically created assets or campaign-level broad match was active.
  3. DSA preparation: Dynamic Search Ads remain active and need a plan before February 2027.
  4. New campaign: AI Max was selected during setup and requires a pre-launch control review.
  5. Not confirmed: no clear account notice or change record exists.

Do not use a general announcement as proof that one campaign migrated. The campaign settings, change history and account notification should agree.

The Maqtrix Controls-First Migration Audit

The Controls-First Migration Audit is a Maqtrix operating framework, not a Google requirement. It asks one question before a feature expands targeting or creative: Can the business see, govern and measure what the system is changing?

A control passes when it has a clear setting, an owner, evidence and a response rule.

1. Preserve the before-state

Before changing a campaign, export or capture campaign and ad-group settings, budgets, bid strategy, locations, brands, keywords, negative keywords, search terms, ads, pinned assets, landing-page rules, conversion actions and recent lead-quality outcomes. Record the date, account, campaign and person making the change.

A screenshot alone is easy to misread later. Pair it with an export and a short change note. This record provides comparison evidence; it does not guarantee that every legacy configuration can be restored after migration.

2. Confirm the conversion signal

Automated bidding learns from the outcomes supplied to it. A form submission called a conversion does not reveal whether the enquiry was qualified, inside the service area, reachable or commercially viable.

  • Confirm that a real person completed a meaningful action.
  • Test whether thank-you-page reloads create duplicates.
  • Separate shallow actions such as page views from primary lead goals.
  • Connect enquiries to qualified leads, opportunities or completed sales where consent and data quality allow it.
  • Compare cost per qualified lead with the platform's cost per reported conversion.

A falling cost per form submission can hide a rising cost per qualified lead. Keep both views.

3. Protect location intent

Service businesses often have a hard operating boundary. An accountant may serve Metro Vancouver remotely, while a home-service contractor may dispatch only within a defined radius.

Google describes location targeting as a best-effort system based on several signals and does not promise 100% accuracy. AI Max also offers locations of interest at the ad-group level. According to Google Ads location guidance, this setting can use the place named or implied in the search while still requiring the user to meet campaign-level geographic criteria.

  • State the physical or remote service boundary.
  • Distinguish presence from interest where that changes eligibility.
  • Add only locations the business can serve.
  • Review excluded areas and out-of-area enquiries.
  • Compare matched location intent with the landing-page promise.
  • Assign an owner to review geographic lead quality weekly.

4. Review query and brand boundaries

AI Max search-term matching can extend beyond existing keywords by using signals from keywords, ads, assets and landing pages. Group recent search terms by service intent, research intent, employment intent, support intent and irrelevant intent. Preserve negative keywords with a documented reason.

Google provides brand inclusions and exclusions, but those controls can limit reach. Use them for a specific business reason. Decide whether branded and non-branded demand should remain separately measurable, then review the AI Max source and match-type reporting.

5. Set landing-page boundaries

Final URL expansion allows Google to choose another relevant page on the same domain. That may help when the website has strong, current pages for each service. It can create a poor journey when the site contains old offers, hiring pages, unsupported locations or articles that cannot complete the intended action.

Eligible destinationsReview before useExclude
Current service pages
Accurate local pages
Working contact or booking paths
Pages with reliable conversion tracking
Educational resources
Older offers
Pages with a different conversion path
Broad industry or location pages
Privacy and terms pages
Careers and recruitment pages
Thank-you pages
Expired promotions
Staging or account pages

Google warns that tracking templates can produce incorrect or broken dynamic landing pages when they are incompatible with final URL expansion. Test representative URLs and parameters before launch.

Check pinned responsive-search-ad assets

Google's AI Max setup documentation says pinned responsive-search-ad assets may not be used when final URL expansion or URL inclusions select a different page. A business that relies on pinned wording for a regulated claim, mandatory qualifier or tightly controlled offer should treat this as a launch decision.

6. Govern generated text

Text customization can use existing ads, website copy, keywords and assets to produce new text. Audit the source material for outdated prices, claims without evidence, old services, changed hours, unsupported areas and language that overstates speed, savings or availability.

Google's text-guidelines documentation describes term exclusions and messaging restrictions for AI-generated assets. The feature is labelled beta and can have limitations.

  • Never state a price that is not present in an approved offer.
  • Do not claim guaranteed savings or guaranteed results.
  • Do not say “24/7” unless that service is staffed at all times.
  • Use the approved business name and spelling.
  • Do not mention cities outside the service area.
  • Avoid superlatives that cannot be substantiated.

Review generated assets in the asset reports and remove inaccurate material promptly. A platform control supports governance; it does not replace human approval.

7. Check budget and operating capacity

AI Max can reach additional queries. Record daily budget, actual spend, target CPA or ROAS, cost per qualified lead, expected contribution from completed work, lead-to-sale rate and the maximum additional volume the team can handle.

Google Ads Help notes that AI Max will show an alert when a campaign is limited by budget. The alert is not an automatic instruction to spend more. A budget decision should follow unit economics, lead quality and capacity. The Ad Budget Calculator can help frame a test budget, but it still depends on real conversion and margin data.

8. Test deliberately and keep decision evidence

Google offers AI Max experiments that divide traffic and budget within an existing campaign. The treatment can activate search-term matching, text customization and final URL expansion by default, while certain controls can be adjusted. Eligibility and available controls vary.

  • Write one decision the test must support.
  • Choose the primary business outcome.
  • Define lead-quality and location guardrails.
  • Document every feature enabled in the trial.
  • Check whether brand controls affect both arms.
  • Set a review date and an owner.
  • State what would justify applying, refining or ending the test.

Google says advertisers should generally wait at least two weeks after enabling AI Max before making performance adjustments so the campaign can learn. Critical problems require immediate action: fix broken URLs, false claims, policy issues and clearly unserviceable traffic as soon as they appear.

The 30-day AI Max monitoring sheet

Use one row per campaign and preserve notes from every review.

Review pointCheckEvidenceDecision
Before changeSettings, spend, terms, URLs, assets, conversions and lead qualityExport, screenshots, CRM baseline and ownerApprove, revise or defer
Launch dayFeature states, tracking, brands, geography and landing-page pathsChange history and live click testsContinue or correct
Days 1–3Broken URLs, copy errors, spend spikes and out-of-area trafficSearch-term, asset and landing-page reportsFix critical issues
Day 7Early query themes, destinations, spend and enquiry qualityQuery categories and lead notesRecord; avoid reacting to noise
Day 14Conversion mix, qualified-lead cost, location quality and assetsTwo-week campaign and CRM comparisonHold, narrow or continue
Day 21Repeat patterns and operating effectQualified leads, booked work and capacityPrepare final decision
Day 30Commercial result and control complianceBaseline comparison, limits and sign-offApply, refine, pause or extend

Track spend, reported conversions, cost per conversion, valid and qualified leads, cost per qualified lead, out-of-area rate, booked appointments, sales, search terms by intent, AI-selected landing pages, removed generated assets and service-capacity incidents.

Small accounts may need longer than 30 days to collect a useful sample. A calendar deadline does not create statistical confidence. Record sample size and uncertainty with the decision.

Two service-business examples

Accounting firm

An accounting firm serves Surrey and Metro Vancouver and wants more advisory engagements. Its website also contains tax-season articles, employment pages and an old promotion.

Before allowing final URL expansion, the firm approves current advisory and service pages, reviews relevant articles and excludes hiring, old-offer and confirmation pages. Text restrictions prohibit guaranteed savings and unapproved pricing. The campaign records booked consultations and qualified opportunities in addition to submitted forms.

The test succeeds only if qualified advisory demand improves without producing misleading claims or sending visitors to the wrong service.

Home-service company

A contractor dispatches within a fixed area and has limited same-day capacity. Search-term expansion could find new demand, but out-of-area calls and inaccurate “emergency” language would waste budget and damage trust.

The company documents its service boundary, reviews location intent weekly, excludes unsupported areas and permits urgency claims only where the operating team can fulfil them. It monitors answered calls, valid service requests, booked jobs and contribution margin.

The test succeeds only when the additional work is serviceable and profitable.

How to make the 30-day decision

FindingRecommended response
Qualified demand improves, economics remain sound and no control failures appearConsider applying or expanding carefully
Reported conversions rise but qualified-lead rate fallsRepair the conversion signal and narrow query, location or destination scope
New queries are relevant but landing pages are inconsistentRefine URL inclusions, exclusions and page content
Generated assets repeatedly violate approved claimsStrengthen guidelines, remove assets or turn off text customization
Out-of-area demand increasesRecheck campaign geography, locations of interest and negative controls
Spend grows faster than service capacityLimit budget or scope until operations can respond well
Sample is too small to distinguish a patternExtend the observation period and state the uncertainty
Broken destinations, false claims or material policy risks appearCorrect or pause immediately

AI Max should earn a wider role through evidence. The setting itself is not the strategy.

Frequently asked questions

Is AI Max a new Google Ads campaign type?

No. Google describes AI Max as an optimization layer within existing Search campaigns. It brings search-term matching, text customization, final URL expansion and related controls into the Search campaign setup.

Are Dynamic Search Ads automatically upgrading in September 2026?

Google revised the schedule. Eligible campaigns using automatically created assets or the campaign-level broad match setting began transitioning in September 2026. Google says the Dynamic Search Ads sunset and automatic transition will begin in February 2027.

Can AI Max be turned off?

Google Ads allows advertisers to disable or modify AI Max features. Existing settings and the effect of an earlier migration can differ by campaign, so capture the original configuration and do not assume disabling AI Max will recreate every legacy DSA structure.

Should a local service business use final URL expansion?

Use it when the website has accurate, current, conversion-ready pages and the campaign has deliberate URL boundaries. Exclude pages that cannot fulfil the search intent, contain outdated information or should never receive paid traffic.

Does AI Max guarantee more leads?

No. Google has published an average performance finding from its internal non-retail advertiser data, but results vary by account. Measure qualified leads, sales and serviceability alongside the platform's reported conversions.

How long should an AI Max test run?

Google recommends waiting at least two weeks after enabling AI Max before making performance changes. The useful test length depends on conversion volume, sales cycle and lead quality. Low-volume service businesses may need more than 30 days.

What is the difference between AI Max and Performance Max?

AI Max is a collection of features within Search campaigns. Performance Max is a different campaign type that can serve across several Google properties. The setup, reporting and controls are not interchangeable.

Sources and further reading

Product facts come from Google. Performance statistics are vendor findings and should be interpreted as such. Sources were checked September 23, 2026.

  1. Google Ads & Commerce Blog. We're upgrading Dynamic Search Ads to AI Max. Published April 15, 2026; updated June 11, 2026.
  2. Google Ads Help. How AI Max for Search campaigns works.
  3. Google Ads Help. Set up AI Max for Search campaigns.
  4. Google Ads Help. About reporting in AI Max for Search campaigns.
  5. Google Ads Help. About AI Max experiments.
  6. Google Ads Help. Target ads to geographic locations.
  7. Google Ads Help. Use text guidelines with Performance Max and Search campaigns.
  8. Google Ads & Commerce Blog. Make AI Max work for your business with new testing and planning tools. Published August 20, 2026.
  9. Google Ads & Commerce Blog. We're bringing AI Brief to more languages and adding a new AI Max reporting feature. Published September 23, 2026. AI Brief is described as a closed beta; additional reporting availability details are still forthcoming.