Google Just Made Saying No Optional in Search Ads — Porter Explained Why That’s the End of Strategy

On September 1, Google began forcibly migrating Search campaigns still running Automatically Created Assets or campaign-level broad match into AI Max. This isn’t an opt-in beta — it’s a reclassification advertisers didn’t request and can’t decline.

Google was cautious enough about the disruption to push the Dynamic Search Ads portion of the migration to February 2027, after advertisers objected to absorbing the change during Q4. But for everyone else, the switch is already flipped.

In the industry’s own State of PPC 2026 survey, 53% of advertisers say Google Ads is harder to manage than it was two years ago, and 62% say the platform now makes decisions they cannot manually override. The trade coverage has framed this as a performance story — Google’s own data shows roughly 7% more conversions from accounts running the full AI Max feature suite.

Most CMOs signing off on the migration are treating that 7% as the whole conversation. It isn’t. The real story is what got removed to produce that lift, and almost nobody approving the change is asking about it.

Porter’s Warning From 1996: The Essence of Strategy Is Choosing What Not to Do

Michael Porter answered this question three decades before AI Max existed. His argument in “What Is Strategy?” is that operational effectiveness — doing similar things better than rivals, on the same tools, toward the same benchmarks — is not strategy, and never produces durable advantage, because every competitor eventually gets access to the same improvement. A defensible position is built on deliberate trade-offs.

A company’s clearest strategic signal isn’t what it does, it’s what it refuses to do. Broad match at the campaign level, automatic search-term matching, algorithmic text customization — these are Google’s operational language for “we will decide which queries you show up for, and you will not be consulted in advance.” A CMO who lets that decision default to the platform hasn’t automated a tactic. They’ve deleted the refusal that used to separate their account from every competitor now running the identical AI Max stack.

When an entire category is optimized by the same model, on the same signals, toward the same lift, that’s not an edge — it’s the convergence Porter spent his career warning boards about.

AI Max Doesn’t Optimize Your Account. It Removes Your Ability to Refuse.

The 7% conversion lift Google is citing describes an average across every advertiser who adopts the feature suite — which by definition cannot be a source of competitive advantage once your direct competitors adopt the same suite next quarter, because Google is selling all of you the identical capability at the same time.

What can still be a source of advantage is the layer sitting above the platform: which queries, audiences, and outcomes a business deliberately excludes because they don’t fit its position, regardless of what the algorithm predicts they’d convert at. Almost no marketing organization has formalized that layer as policy.

Most have a media buyer watching blended CPA and a platform rep recommending “let AI Max decide” — and 62% of advertisers already admit, in their own survey responses, that they can no longer manually override the result.

That isn’t a workflow gap. It’s a vacancy where a strategic decision used to sit, and Google is glad to fill it by default.

The Targeting Call at Bolt Goodly That Cost Us Revenue on Purpose

I made this exact trade-off deliberately, years before AI Max existed. A financial-services client at Bolt Goodly wanted broad match turned on across the account because the weekly report looked better — more impressions, more clicks, a lower blended CPC.

I argued for keeping a strict negative-keyword list and phrase-match discipline on their highest-margin service line, and we refused an estimated $30,000 a month in additional traffic that the account data said would convert at a materially lower rate and burden the sales team with weaker leads.

The weekly dashboard looked worse for two straight quarters. The sales team’s win rate on marketing-sourced leads rose 40%, and that service line became the agency’s highest-margin engagement within a year — because we refused traffic instead of chasing volume.

That refusal was the strategy. Run on autopilot, AI Max would have overridden it in week one, and the dashboard would have looked great while the business got worse.

What Every CMO Approving AI Max Should Actually Be Deciding

This is the trade every CMO signing the migration off this quarter is making without naming it as a trade at all. A 7% average lift is not a competitive advantage when every rival gets it simultaneously — it’s table stakes, arriving on a schedule Google controls, not you.

The advantage question was never “should we adopt AI Max,” since for most advertisers that’s no longer a live choice. The question is which queries, audiences, and use cases your business commits, in writing, to refusing regardless of what the algorithm predicts — and whether anyone in the organization currently owns that decision, or whether it’s just drifting to whatever Google’s default hasn’t gotten around to overriding yet.

Boards evaluating AI marketing spend this fall should be asking a sharper version of the usual question: not “is this improving performance,” but “if every competitor in our category adopted this exact capability today, what would still make a customer choose us?” If the honest answer is “nothing, but our numbers are up 7%,” the organization hasn’t found efficiency. It’s found the convergence Porter described in 1996, wearing a performance-marketing dashboard.

The Question for the Next Board Meeting

Porter’s answer was never “automate less.” It was “know precisely what you’re refusing, and don’t let a machine make that refusal for you by default.” The CMOs with a defensible position in this category two years from now won’t be the ones who resisted AI Max — most won’t have had the choice, since Google isn’t asking permission.

They’ll be the ones who decided, deliberately and on the record, exactly which queries, audiences, and outcomes their account is built to say no to, and who treated that refusal as the strategy rather than as revenue left on the table.

Before the last legacy campaign finishes migrating, there’s one question the platform cannot answer on your behalf: what is your account still refusing to do — and is that a decision your business made, or just something Google hasn’t automated away yet?