Google Ads AI Max Auto-Migration (September 1–30, 2026): How Solo Founders Can Tell the Upgrade From Everything Else in Their Account

Beginning September 1, 2026, Google Ads will automatically upgrade Search campaigns using campaign-level broad match or standalone Automatically Created Assets to AI Max, and the rollout runs gradually through September 30 rather than flipping on a single day. Google closed the door on the old setup on August 3, when it stopped allowing new campaign-level broad match configurations and legacy ACA campaigns to be created in the Google Ads interface, Ads Editor and API, according to Search Engine Land.

For an advertiser with a dedicated PPC analyst, this is a calendar entry. For a founder who checks the account on Sunday nights, it is a measurement problem: something in your Search performance will move in September, and you will not know whether AI Max did it, whether Q4 auction pressure did it, or whether your landing page has been broken since Tuesday.

What is actually changing in Google Ads on September 1, 2026?

Two legacy features are being retired into AI Max: the campaign-level broad match setting and standalone Automatically Created Assets. ACA campaigns move with search term matching and text customization enabled by default; campaign-level broad match campaigns move with search term matching enabled by default. Google says campaigns migrate in place using equivalent AI Max settings, and that brand inclusions and exclusions carry over.

Search Engine Roundtable’s Barry Schwartz published the advertiser email on August 7, sourced from a screenshot posted by Menachem Ani. The wording is worth reading literally: campaigns “will automatically be upgraded.” No opt-out is described, only a window. Check whether you are affected rather than assuming — ACA was on by default in plenty of accounts someone else built and handed over.

Why is a 30-day rollout window harder to handle than a single deadline?

Because you lose the before-and-after line. A migration rolling out “gradually throughout the month” means your campaign changes on an unannounced date, and every comparison you make afterwards blends pre- and post-migration days.

Staggered rollouts are standard engineering practice and genuinely reduce the blast radius of a bad release. But the benefit is asymmetric: Google gets a controlled ramp, the advertiser gets an unlabelled change inside their own data. An enterprise team absorbs that with an analyst. A solo founder absorbs it by guessing.

How does search term matching differ from campaign-level broad match?

Both expand the queries you show for. The difference is what they expand from. Broad match starts with a keyword you typed and reasons outward. AI Max search term matching also reads your landing pages and ad assets, generating query coverage from that material.

That change in seed material is the whole mechanism, and it predicts cases nobody has written about yet. Any text on your site describing something you can do — but do not want to sell — becomes eligible targeting input. A roofing company whose footer lists gutter cleaning as a courtesy now has gutter cleaning in the seed set. A B2B SaaS whose pricing page mentions a free plan it deprioritized last quarter now has free-tier shoppers in it.

Negative keywords still apply, and you should use them. But a blocklist only works against terms you thought of in advance, and when the seed set expands from “keywords I chose” to “everything my site says about me,” the terms you must anticipate grow faster than the hours you have to anticipate them.

Does “migrated in place with equivalent settings” mean performance stays the same?

No. Equivalent settings and equivalent outcomes are different claims, and only the first is being made. Configuration equivalence means the toggles land in the closest matching position; it says nothing about the matching engine underneath them, which is the part being replaced. Google is being accurate. The industry is the one rounding it up to “nothing will change.”

Here is the shape of the stakes at a small-account budget. Take a local service business spending $2,000 a month on Search at a $60 cost per lead — roughly 33 leads. Suppose one-fifth of spend, $400, shifts to a broader query pool converting at half the rate. That $400 now returns about three leads instead of about seven. The month lands near 30 leads and blended cost per lead rises to roughly $67, an 11% increase, with no setting visibly changed. That is illustrative math, not a forecast — but 11% is exactly the size of move that is easy to write off as a bad month and expensive to write off twelve times.

Why is positioning the right lens here, not just CPA?

The instinct is to evaluate this through direct response: did my cost per acquisition go up? Fair question, wrong first question.

Positioning — the school associated with Al Ries and Jack Trout — holds that you do not own a product, you own a slot in the prospect’s mind relative to the alternatives they are already comparing. Your keyword list has always been the cheapest available statement of which slot you were claiming: every query you bid on declared a category you competed in and a set of competitors you accepted being priced against.

Search term matching moves that boundary from something you assert to something the system infers from your site copy. A specialist can get quietly reclassified as a generalist, and a generalist competes against everyone. That damage does not show up as a CPA spike in month one. It shows up as conversion volume holding steady while lead quality drifts down — the failure a CPA-only dashboard is blind to.

How should a small advertiser diagnose a September performance change?

Run this before September 1. Every step is executable in one sitting by one person.

1. Confirm you are in scope. Check each Search campaign’s settings for the campaign-level broad match toggle, then the assets tab for automatically created assets. If neither is present, September is a normal month.
2. Export a 60-day baseline today. Search terms report plus impressions, CTR, conversions, cost per conversion and search impression share for July 1 through August 31. You cannot reconstruct this later.
3. Write down three numbers. Baseline CPA, conversion rate, and revenue or value per conversion. Track only the first two and you will not see quality drift.
4. Read your landing pages as targeting input. List every service or capability mentioned on the pages your ads point to. Anything you do not want to sell is a candidate negative. Highest-value hour in the process.
5. Build negatives from your positioning, not your search terms report. Search terms tell you what already happened; positioning tells you what should never happen. Add adjacent categories, free-tier seekers, DIY intent and job seekers now.
6. Date the migration yourself. Check Change History weekly through September and note the day the AI Max upgrade appears on each campaign.
7. Freeze bid and budget changes for 14 days after that date. Learning-phase noise and migration effects are indistinguishable if you edit on top of both. Fix broken things; do not optimize.
8. If you run multiple accounts, batch it. Eight clients means eight baselines and eight negative lists — roughly 40 minutes each, so a full day. Schedule it in August.

Three symptoms, three readings. Impressions up with CTR down and CPA flat or rising points at query expansion. Impressions flat with CPA up points at auction pressure rather than migration; Q4 competition starts building in September across most retail-adjacent verticals. Conversions holding steady while value per conversion falls is the positioning symptom, and you will only catch it if you wrote down that third number.

What dates come after September 2026?

• August 3, 2026 (passed): new campaign-level broad match and legacy ACA blocked across UI, Ads Editor and API.
• September 1–30, 2026: automatic migration of broad match and ACA campaigns to AI Max, rolling out gradually.
• September 2026: in-account notices begin encouraging voluntary DSA upgrades.
• January 15, 2027: reminders to advertisers still running legacy DSAs.
• February 1–28, 2027: automatic DSA migration; creating new DSA ad groups is permanently removed.
• Approximately September 2027: older Google Ads API versions supporting legacy broad match and ACA entities reach scheduled sunset.

Two of those dates collide with cash-sensitive months. September overlaps the start of Q4 cost-per-click inflation, and the February DSA migration lands in the post-holiday demand trough, when small advertisers have the least headroom to absorb a swing. Plan the DSA work in January.

Frequently asked questions

Can I opt out of the AI Max migration?

Google’s advertiser email describes automatic upgrades, not an opt-out. What you retain is control over the settings inside AI Max after migration: search term matching and text customization can be reviewed, and brand inclusions and exclusions carry over. Plan on configuring the destination rather than avoiding the trip.

Do my negative keywords still work under AI Max?

Yes, negatives continue to apply. The practical limitation is coverage rather than function: search term matching draws on your landing pages and assets as well as your keywords, so the range of queries you need to anticipate widens. Build negatives from the categories you refuse to compete in, not only from terms already in your report.

How will I know the exact day my campaign migrated?

Change History is your best available record. Check it weekly through September, note when the AI Max upgrade appears on each campaign, and annotate that date in whatever reporting you keep. Because the rollout is gradual, campaigns in the same account may migrate on different days.

Does this affect Dynamic Search Ads in September?

No. DSAs are excluded from the September window. Google plans in-account notices in September encouraging voluntary upgrades, reminders on January 15, 2027, then automatic migration between February 1 and 28, 2027, after which new DSA ad groups cannot be created.

What should you do this week?

Export the baseline and read your landing pages as targeting input. Two hours, and they are the difference between diagnosing September and arguing with it. Everything else can wait. The baseline cannot.

The schools of marketing thought we write from are collected at CampaignPress.ai, and the audit workflows built for small accounts live in the Campaign Press tools. Google now decides how your campaigns match. What you are matchable for is still yours.