Address

30 N Gould St Ste N, Sheridan, WY 82801

Phone number

+212 681 53 04 05

Email

contact@skyweb3agency.com

Anyone who has run Google Ads for more than a few years has watched their job change without being consulted. Match types that once meant precision now target “related intent.” A 2023 rebuild made Broad Match genuinely competitive for the first time. Smart Bidding shifted the conversation from keywords to outcomes like ROAS and CPA. And now, with AI Max, keywords are becoming optional in Search campaigns altogether.

One of Google’s first AdWords Evangelists — an employee since 2002 who spent a decade building the keyword-based system PPC has run on ever since — recently put it bluntly: the keyword, as advertisers have known it, is dead. Not as a slogan, but as a technical description of what’s happening under the hood. The core targeting system is being rebuilt even while the legacy interface sticks around. As users shift from typed queries to conversational prompts, a “synthetic keyword” — a machine-generated distillation of intent — is quietly replacing the keyword an advertiser typed into an ad group. The industry isn’t fully there yet, but anyone still defining PPC as “picking the right keywords” is standing on ground that’s already shifting.

What the old contract actually was

For most of Google Ads’ history, keywords functioned like a contract. Advertisers researched relevant terms, structured accounts around them, and wrote ads that spoke directly to that intent. In exchange, Google agreed to show those ads only when a matching query — based on the chosen match type — triggered the auction. Exact meant exact. Phrase meant phrase. Broad was the wildcard advertisers used when they were willing to trade precision for reach.

That arrangement produced something genuinely valuable: diagnosability. When a campaign underperformed, the search terms report showed exactly what was being paid for, line by line. Bad queries got negatived, good queries got promoted, and match type was the main lever PPC managers used with real care. It worked for as long as it did because the underlying technology simply couldn’t do much useful with intent beyond a list of typed keywords.

How the deal came apart

The shift wasn’t a single announcement — it built up over roughly a decade of incremental product decisions. Close variants arrived first, with Exact Match quietly absorbing misspellings, then plurals, then function-word variations, until “exact” no longer meant what the label implied. Smart Bidding moved the center of gravity next: once bids were set against conversion probability rather than keyword bids directly, which keyword triggered an auction mattered less than whether that user was likely to convert. Match type became less a targeting rule and more a throttle on how aggressively the system could explore new queries.

The 2023 Broad Match overhaul changed the narrative further. Google invested real engineering into making Broad semantically intelligent and reported roughly 25% more conversions in Target CPA campaigns using it — a reversal for advertisers who’d spent fifteen years treating Broad as a budget leak.

AI Max is where the synthetic keyword becomes explicit. Feed Google a URL, creative assets and business data, and the system infers intent on its own — keywords become optional from the advertiser’s side, though the auction itself still runs on a keyword substrate underneath. What’s changed is who’s selecting those keywords and how visible that selection is.

This isn’t unique to Google, either. Advertising firm Optmyzr has begun placing ads on ChatGPT, where keywords are optional from day one — advertisers supply business signals, and the system matches ads to the shape of a user’s question rather than a phrase declared in advance. When the company that defined keyword advertising and the company reinventing search both converge on keyword-optional matching, that’s a signal intent has outgrown the keyword as a targeting unit. Related shifts in bidding automation are worth reviewing — see where to spend your time on Google Ads bidding strategies in 2026.

What gets lost, and what gets gained

This transition isn’t costless. Granular diagnosability is the first casualty — the old playbook of reading the search terms report, finding bad queries and tightening match type only half-works on a keyword-less campaign, because the intent-matching engine underneath is harder to reason about. The craft of account structure is the second loss: tight ad groups and clean branded-versus-non-branded separation were long a proxy for control, and that structure loses strategic value once the system handles more targeting itself. And training suffers, since junior analysts used to build intuition by watching the search terms report for weeks, learning how language and intent drift — one of the discipline’s best teaching tools is quietly disappearing.

Against those losses sit real gains. Coverage improves for queries no keyword list ever caught — zero-click queries, new phrasing, localized slang — because the query space is too large to enumerate manually. Maintenance overhead drops: thousand-line negative lists, endless query audits and single-keyword ad group construction existed largely to bridge the gap between advertiser language and user language, and closing that gap algorithmically frees up time for strategy and creative. And advertisers gain access to signals no manual process can match, since Google’s language-model-driven query understanding sees more of a user’s journey than any keyword list ever will.

What the data shows

Optmyzr’s 2026 Match Type Study, run across nearly 130,000 non-branded campaigns in more than 14,000 accounts totaling roughly $99 million in spend, is the clearest quantitative evidence that practitioners are already adapting to this shift, whether or not they’d describe it that way.

  • Exact Match’s share of non-branded spend has fallen from 37.1% in 2022 to 27.6% today, with most of that decline in the last two years.
  • On branded terms the pattern reverses: Exact Match delivers 6.61x ROAS at a $0.90 CPC, nearly double either alternative. Brand intent is known intent, and Exact still owns it.
  • Phrase Match has become the workhorse, driving 40% of non-branded conversions at a 15.7% conversion rate — well above Exact’s 10.5% and Broad’s 8.5%.
  • Broad Match keeps growing, now representing 38.8% of non-branded spend, the single largest bucket, even though its ROAS still trails the other two.

The pattern is consistent: the industry has been migrating toward looser, intent-driven matching for non-branded queries while keeping tight control where intent is already certain. AI Max simply turns that same dial further. One agency practitioner interviewed alongside this research described a playbook that mirrors the data: start new lead-gen campaigns on manual CPC and Phrase Match, collect a few weeks of quality traffic, promote winners to Exact, then layer Broad Match and Smart Bidding on top once the data is proven — Exact is “too specific” for a brand-new campaign, Broad is “overly aggressive” too early, and Phrase is the sweet middle spot.

Five adjustments worth making now

1. Separate branded and non-branded campaigns cleanly

Sloppy brand separation is the difference between 6.61x ROAS and roughly half that. Build a dedicated branded campaign, lock it to Exact, and stop letting non-branded traffic creep in.

2. Invest in the signals the system actually reads

Landing pages, feed quality and business data aren’t just conversion-rate inputs anymore — they’re the targeting inputs AI Max uses to decide when an ad appears. For lead generation, piping qualified-lead data back to Google matters more than keyword refinement now; roughly 100 leads a month is usually the threshold needed to generate enough qualified conversions for Smart Bidding and AI Max to optimize toward meaningfully. Clean reporting matters just as much here — the discipline covered in how to audit and fix value inflation in Google Ads applies directly, since a bidding system fed bad values chases the wrong signal regardless of match type.

3. Treat negative keywords as the last real lever

Negatives still work in the AI Max era and remain the most powerful tool for telling the system “not that, ever.” Maintain them aggressively and automate additions where possible.

4. Test AI Max where Performance Max would already run, with a hold-out

The mechanics and prerequisites are similar — sufficient conversion volume, clean tracking, and a business the algorithm recognizes. The accounts getting real value are the ones running it against a proper hold-out and measuring incrementality, not the ones that flipped it on and walked away, while keeping keyword-based campaigns live on branded and proven high-intent queries where those still hold an edge. That’s the same discipline worth applying as agentic commerce reshapes Google Ads more broadly — measure before you commit budget.

5. Upgrade the skill set, not just the tactics

The PPC manager of the next few years isn’t a keyword picker — they’re closer to an intent engineer, translating business goals into signals the system can learn from, debugging a semi-black box through query reports and controlled experiments, and explaining what’s actually happening inside an account when the dashboard only shows aggregates.

Control versus leverage

The system built over the past two decades was designed for control. What’s replacing it is designed for leverage. Conflating the two explains a lot of the frustration practitioners feel toward tools that are, by the numbers, performing well. Control meant dictating terms to Google directly; leverage means feeding the engine the right signals and letting the auction execute at a scale no human team can match. The data suggests the industry is already roughly halfway through this transition — the real question isn’t whether to adapt, but how quickly teams can shift from defending the old interface to mastering the inputs the new one requires.

Frequently asked questions

Are keywords completely gone from Google Ads?

No. The auction still runs on a keyword substrate underneath, and match types like Exact and Phrase still matter, especially for branded terms. Advertisers just no longer have to declare every keyword themselves.

Should I stop using Exact Match?

No. Exact still delivers the highest ROAS on branded terms by a wide margin. The shift is toward using Phrase and Broad more for non-branded discovery while keeping Exact for known, high-intent queries.

What should I prioritize instead of building keyword lists?

Landing page quality, feed and asset hygiene, clean conversion tracking, and offline conversion data feeding back into the platform — the signals automated matching systems actually use to decide targeting.

Is AI Max the same as Performance Max?

They share underlying mechanics, but AI Max operates within Search campaigns rather than replacing them. Both need solid conversion volume and clean tracking, and both are worth testing against a hold-out group before committing full budget.

Leave a Reply

Your email address will not be published. Required fields are marked *