What changed
Google is collapsing three standalone Search features into a single campaign-level setting called AI Max. Dynamic Search Ads (DSA), which auto-generated headlines and landing pages from your website crawl, are gone as a campaign type: no new DSAs can be created after September 2026, and existing ones stop serving in February 2027. Automatically created assets (ACA), which let Google write additional headlines and descriptions for existing ads, and campaign-level broad match, which applied broad match to every keyword in a campaign at once, both auto-migrate to AI Max in September 2026. There is no opt-out for that second phase.
AI Max is a campaign-level toggle that ships with three capabilities turned on by default: search term matching (broadens query coverage beyond exact keyword intent), text customization (Gemini generates new headlines and descriptions), and final URL expansion (routes the user to the page Google thinks is most relevant, which may not be your chosen destination URL). You can turn final URL expansion off. You can constrain text customization with guidelines. You can protect brand terms with brand controls. But those guardrails don't configure themselves.
Who it affects
If you're running DSA campaigns, you have the clearest deadline: migrate before September or lose the ability to create anything new. If you're using ACA for headline volume or campaign-level broad match for simplified keyword management, auto-migration will happen regardless. Accounts in competitive verticals, insurance, legal, finance, e-commerce with thin margins, feel this most acutely because query expansion and URL routing are the two levers most likely to move CPA in the wrong direction before you notice.
Small accounts that relied on DSA as a low-maintenance way to capture long-tail queries need a new plan. Large accounts with strict brand guidelines need to audit AI Max's brand controls before September, not after.
Why it matters
The structural shift here is bigger than a feature rename. DSA, ACA, and broad match were three separate levers with separate reporting surfaces and separate controls. AI Max fuses them. That means less granular control by default, reporting you'll need to relearn (search terms report and asset-level data replace DSA query reports), and Google's models making more decisions per impression than before.
The 7% conversion lift Google cites for the full AI Max suite versus search term matching alone is a platform-level average. Averages mask the distribution. Accounts with tight creative, well-structured landing pages, and strong Quality Scores are more likely to land above that line. Accounts with sprawling URL structures or loose brand guidelines are more likely to land below it, and may not realize it for weeks because attribution lag (the gap between click and reported conversion) obscures early performance signals.
Final URL expansion is the highest-risk component for most advertisers. Google's algorithm may route a user searching for a specific product to your homepage, or a blog post, because it scores that URL as "most relevant." In practice this can drop conversion rate while holding click volume, producing a CPA increase that looks fine at the campaign level until you cut by landing page.
The play
-
Audit DSA campaigns now. List every DSA campaign, its historical query coverage, and the ad groups you'd need to replicate in standard format. Use Google's upgrade tool to port settings, but verify the output, auto-ported structure often collapses tightly themed DSA targets into a single ad group. The metric to watch is impression share post-migration; a drop above 15% signals coverage gaps worth fixing manually.
-
Migrate one ACA or broad-match campaign voluntarily before August. The goal is to configure brand controls, URL expansion rules, and text guidelines in a controlled environment before September forces the switch. Measure CPA against a 30-day pre-migration baseline. A 10% CPA increase in the first two weeks is worth tolerating to keep configuration control; an increase that persists into week four is a signal to tighten text guidelines or pull URL expansion off.
-
Turn off final URL expansion first, then test it in. The default-on behavior is the biggest source of unexpected CPA movement. Disable it at launch, establish a baseline, then re-enable it on a subset of campaigns and measure conversion rate by landing page. The bar: conversion rate on expanded URLs should match your primary destination within 15% before you let it run broadly.
-
Build a brand control list before you touch anything else. Competitor brand terms, your own branded keywords, and any category terms where you don't want Gemini writing copy should all be listed in brand controls before text customization goes live. Audit the search terms report weekly for the first month.
Watch-outs
Auto-migration in September will not prompt you to set up brand controls. If you haven't done it by then, AI Max defaults will apply, and competitor brand terms can surface in your query mix. That's recoverable but messy and potentially expensive depending on your vertical's CPCs.
Don't migrate everything at once. Stagger by campaign priority. If a campaign drives 60% of your revenue, it's the last one you migrate, not the first.
The asset-level reporting in AI Max is better than ACA's legacy reporting, but it requires a different read than classic ad-level data. Gemini-generated assets don't get the same historical quality score context that human-written assets carry. Give the system at least two to three weeks of data before making optimization calls on individual assets.
The WhyItWon angle
AI Max introduces three new generative surfaces, query matching, copy generation, and URL routing, operating simultaneously inside a single campaign. For most accounts, that means the question "what creative will actually win in this new environment?" is genuinely harder to answer than it was under DSA or static ACA. More of the variation is happening inside Google's model, less of it is visible ahead of spend. That's exactly the problem WhyItWon is built for: reading what's already working across your ads and your competitors', and scoring the next creative before you commit budget to finding out the hard way.
Google Ads
