How Should B2B Bid When AI Overviews Answer the Query?

Classify queries by intent and AI visibility to decide where to bid, influence the answer, or shift budget to demand generation. AI Overviews reshaped the search results page, yet most paid search playbooks still run as if nothing happened. When Google answers a question directly on the results page, the value of a paid click varies across your keywords.

What Happened

Ana Kostic laid out a three-bucket framework for paid search in the AI Overviews era: Fight, Influence, and Generate Demand. Fight covers bottom-of-funnel commercial queries where a click still converts. Influence covers informational queries the Overview resolves, where the goal shifts to being cited inside the answer. Generate Demand covers queries so zero-click that paid search alone cannot rebuild the journey. One bid lever cannot solve all three.

The Pattern

• SE Ranking (July 2026): ads appeared on 54% of commercial AI Mode queries priced at $10+ CPC, versus 24% under $2.
• About 7 in 10 AI Mode answers containing an ad showed two advertisers in the block.
• Shashi Thakur at Google Marketing Live EMEA 2026: the best ads must be answers, not interruptions.

Why This Matters for B2B Marketing Leaders

If you sell HR Tech or FinTech, most of your pipeline sits in queries that never had strong click economics to begin with. Category education, feature comparisons, and buyer research now resolve inside the Overview. Your team is likely still bidding on those terms with the same CPA target you used in 2023, watching cost per lead climb while the AI answer quietly educates your buyer without you in it. Meanwhile, competitor shortlists and demo-intent terms are getting more expensive because paid inventory concentrates where commercial weight already lives. You need three separate measurement models, not one, and you need to know which bucket every keyword belongs in before your next QBR.

The Starr Conspiracy's Take

The Fight, Influence, Generate Demand split maps almost perfectly onto how we think about demand states in the AI buyer's journey. In enterprise HR Tech, maybe 15% of your keyword footprint is Fight. The rest is Influence or Demand Generation, and most marketing teams are still paying Fight prices for Influence queries that resolve inside the Overview. The strategic move is not smarter bidding. It is reallocating spend toward the surfaces that actually shape the answer: original research, expert content, and category presence dense enough that the model names you unprompted.

What to Watch Next

Expect Google to expand ad inventory inside AI Mode through 2026, likely with new formats that blur ads and citations. Watch your Absolute Top impression share weekly. If it drops without a bid change, the AI block is pushing you down and your Fight bucket needs to shrink.

Related Questions

How do you know if a keyword belongs in Fight, Influence, or Generate Demand?

Run the query in incognito. If an AI Overview appears and answers the question completely, it is Influence or Demand Generation. If a shopping or comparison SERP still dominates and ads sit above the fold, it is Fight. Layer intent modifiers like pricing, demo, or partner to confirm commercial weight.

Should FinTech brands cut paid search budget entirely on informational queries?

Not cut, redirect. Move that spend into content and PR designed to be cited by the models. Our take on answer engine optimization for B2B covers which content formats get pulled into AI answers most reliably in regulated categories.

What metric replaces CPA for Influence-bucket keywords?

Citation share and brand mention rate inside AI answers. Track how often your brand appears in Overview responses for your priority query set, and measure branded search lift as the downstream signal. CPA on those terms will look terrible because the click is not the conversion event anymore.