Perplexity vs AI Max in 2026: How the Ad-Model Race Reversed

Why Perplexity killed its ad model in February 2026

The original bet was simple and, briefly, correct. Perplexity would sell Sponsored Follow-Up Questions in the "Related Questions" module (which it says accounts for roughly 40% of platform queries), price them at a CPM north of $50, and offer category exclusivity across fifteen verticals — technology, health, finance, food and the rest. Brands like Whole Foods and Indeed ran pilots. The pitch was "AI-native search ads, with editorial distance: Perplexity writes the answer, the brand just sponsors the question." That was the whole value prop.

It died for three reasons. First, user trust became the product. Perplexity's leadership concluded that the "answer engine" positioning and the ad product were incompatible — every sponsored question lowered perceived objectivity, and objectivity was what Pro subscribers were paying $20/mo for. Second, the subscription ladder started working: Pro at $20, Max at $200 (early Comet access, Email Assistant, premium models) and Comet Plus at $5 turned a cost centre into a revenue line without ad ops. Third, the ARR growth made the ad revenue rounding-error small. When ARR jumped ~50% in a single month to over $450M in early 2026, the ad program was a distraction, not a strategy.

The tell was Comet. Perplexity has said publicly that the browser is designed in part to collect behavioural data for ad targeting — and yet the moment it had the data, it turned off the ads. Read that either as a principled decision or as a long game: the data is still being collected, the ad product can come back, and when it does it will be targeted at a different layer entirely. Either way, for now, there is no Perplexity ad channel to buy.

Google AI Max: out of beta, and coming for your DSA campaigns

While Perplexity exited, Google finished. AI Max for Search — announced at Google Marketing Live 2025 and in beta for most of the last year — was declared generally available on 15 April 2026. It is not a new campaign type. It is a bundle of three features grafted into the Search campaign you already run: search term matching (broader, LLM-mediated matching against query intent rather than keyword form), text customisation (Google rewrites your headlines and descriptions on the fly), and final URL expansion (Google can route the ad to the most relevant page on your domain, not just the one you specified).

The two numbers that matter. First, Google's own benchmark: campaigns using the full AI Max feature suite see 7% more conversions or conversion value at similar CPA/ROAS versus search term matching alone. Second, the deadline. Starting in September 2026, every Search campaign still using Dynamic Search Ads, Automatically Created Assets or campaign-level broad match will auto-migrate to AI Max in September 2026. You will not be able to create new DSA campaigns via the Google Ads UI, Editor, or API after that date.

Functionally this is Performance Max logic applied to Search — more automation, fewer levers, a bigger role for creative variety and landing-page depth, a smaller role for keyword lists. It is also the clearest signal yet that Google intends to meet AI-assistant search head-on by making its own auction more generative, not by launching a separate product.

Sponsored Questions vs AI Max: what actually competes now

The original version of this post compared two ad products. In April 2026, you're really comparing two business models.

Perplexity is now a pure subscription play for consumers and a revenue-share play for publishers via Comet Plus. The $42.5M publisher pool, the 80% Comet Plus revenue share, and the refusal to put sponsored content inside answers are all facets of the same bet: if AI search replaces traditional search for the people who'll pay for it, you don't need ads — you need a subscription flywheel and aligned content supply. We covered the publisher side of this in our earlier piece on Perplexity's revenue-share model.

Google AI Max is the opposite bet: the auction still works, advertisers still want reach, so keep the auction and let the model do the matching, writing and routing. That shifts the advertiser's job from "write tight ad copy against exact-match keywords" to "feed the model good creative, good landing pages, and good conversion signal." The winners will be the accounts with the deepest first-party data and the broadest, most modular creative libraries. Which, as we've argued in how AI Max transforms blog ads and AI Max: the digital marketing revolution, is a content problem dressed up as a media problem.

What this means for marketers

Three practical calls to make this quarter.

1. Stop planning around Perplexity ads. There is no buy-side channel. If you want visibility inside Perplexity answers, that is an SEO / content problem now — publish authoritative content, get cited, structure pages so they're extractable. Comet Plus is a publisher-revenue play, not a media buy.
2. Audit your Search campaigns before September. Every DSA, every campaign-level broad match, every reliance on Automatically Created Assets is going to become AI Max whether you prepare or not. The ones that prepare — with clean asset libraries, brand inclusions/exclusions, URL contains/excludes, and properly-labelled landing pages — keep control. The ones that don't will spend their Q4 unpicking messy auto-migrations.
3. Invest in the creative and landing-page layer. AI Max rewards variety at the asset level. Ten landing pages beat one. Twenty headlines beat three. First-party conversion data beats aggregate ROAS. This is the opposite of how most SMB Search accounts are structured — and it's where the 7% lift lives.

The bigger pattern. Both Perplexity and Google have arrived at the same conclusion from opposite ends: humans writing individual ads, targeting individual keywords, for individual queries, is over. Perplexity removed the inventory. Google removed the inputs. What's left — on both platforms — is operations at content velocity. This is where Anjin fits.

Anjin: the Marketing Operating System for the post-ad-auction era

Anjin is the Marketing Operating System built for exactly this moment. AI Max changes what wins in Google Search. Perplexity's subscription pivot changes what wins in AI search. In both worlds, the unit of competitive advantage is the same: how fast can your marketing team ship high-quality content, landing pages, creative variants, and conversion experiences against a changing model?

That is what Anjin is built for. One workspace where briefs become landing pages, landing pages become ad assets, assets become AI Max feeds, and performance data flows back into the brief. Not another point tool bolted onto your stack. The operating layer that makes the stack move at model speed.

For AI Max specifically, that means a 20-asset library instead of a 3-asset one, without a new hire. For AI search, it means structured, citable content published on a cadence the model notices. For agencies and in-house teams, it means the volume the new ad products demand — without burning another year waiting for your current tooling to catch up.

Sources: Google Ads Blog, Google Business Announcements, Google Ads Help (AI Max), PYMNTS, Perplexity Hub, Comet Plus, AdExchanger, Sacra.