Google Is Now Spending Your Ad Budget Inside AI Answers. Here's How to See What It's Doing.

At Google I/O last week, most of the headlines went to the redesigned search box and the new AI agents. Buried underneath all of that was a change that hits dealers in a more direct place: your ad budget.

Google is now placing ads inside its AI answers, both the AI Overviews that sit at the top of a results page and the new conversational AI Mode. And the part most dealers haven't caught yet is that there's no opt-in. If you're running Search, Shopping, or Performance Max campaigns, your ads are already eligible to serve in those placements. Google made that decision for you.

I'm a dealer advocate first, so let me be clear about where I'm coming from. This isn't a "Google is evil" post. It's a "know what you're paying for" post. Google monetizes new surfaces fast, and the dealers who get hurt are the ones who don't realize the rules changed until the spend has already gone out the door.

The real problem isn't the ads. It's that you can't see them.

Here's the catch that matters. Ads in AI answers are not a campaign type you can turn on, turn off, or report on by itself. Google treats them as a placement, and all of that activity gets blended into your standard search reporting. Google has confirmed there's no separate line item, and as of now the company will only say that segmented reporting is something it's "thinking about." No timeline.

So your money is flowing into a brand-new place, and your reporting can't tell you what it earned there. For an industry that already struggles to tie marketing spend back to a sold unit, that's a problem worth taking seriously.

The wrinkle that's specific to car dealers

Most AI answers don't fire on the bottom-of-funnel searches that put someone on your lot this weekend. They fire on research-phase questions. Think "most reliable used truck under $30k" or "is a certified pre-owned worth it," not "Silverado dealer near me."

That distinction changes how you should think about this entirely.

An ad showing up inside a research-phase AI answer is not a bad buy. It's a top-of-funnel buy. It's reaching a shopper who's three weeks from a purchase, not three minutes. The danger is not that the placement is wasteful. The danger is that this top-of-funnel spend is now getting blended into the same reporting bucket as your bottom-of-funnel inventory ads, and measured the same way.

If you judge a research-phase impression by your cost-per-sold-unit number, you'll conclude it failed, when it was never supposed to be a closing placement in the first place. That's how good top-of-funnel spend gets cut for the wrong reason, which is exactly the kind of mistake that hands the research-phase shopper to a competitor before they ever start searching inventory.

Three things you can actually check this week

You can't isolate AI-answer ad performance directly. But you are not flying blind, and none of this requires buying anything from a vendor. Here are three moves that tell you what's really going on.

1. Segment your search campaigns by Network

In your Google Ads account, open a search campaign, click the Segment icon, and choose Network (with search partners).

You won't see an "AI Overviews" line. That doesn't exist yet. But this view does separate your core Google Search performance from the Search Partners Network, which is a different placement that often runs at a meaningfully higher cost per click and quietly drags on efficiency. While you're in there auditing what the AI rollout is doing, this is worth a look on its own. If Search Partners is costing you far more per click than core Search and not converting, you can turn it off in your campaign network settings.

2. Watch the gap between impressions and clicks

This is the closest thing to a fingerprint for AI placement. When your ad shows up inside an AI answer, a lot of people read the answer without clicking through to you. So the tell is a jump in impressions with flat or falling clicks.

If you pull your search campaign trend since late May and see impressions climbing while clicks stay put, that's a strong signal your ads are increasingly appearing inside AI answers. That's not automatically bad. It can mean more top-of-funnel exposure. But you want to know it's happening so you're not misreading a falling click-through rate as a content or creative failure when it's really a placement shift.

3. Split your campaigns by funnel stage

This is the most important one, and it's the fix for the dealer wrinkle above.

Keep your research-intent keywords in their own campaigns, separate from your in-market and inventory campaigns. Questions and broad, comparison-style terms go in one bucket. "Near me," model-plus-trim, and VDP-style terms go in another.

If AI-answer ads are going to blend into your reporting no matter what (and right now they are), then at least make sure they blend within a campaign whose funnel stage you already understand. When research-phase spend is mixed in with research-phase campaigns, the blended number still means something. When it's mixed in with your closing campaigns, the number lies to you.

The bottom line

2026 is an out-smart year, not an out-spend year. Google built a new place to spend your money and opted you in automatically. That's not a reason to panic, and it's definitely not a reason to start slashing campaigns. It's a reason to get your reporting structured so the spend tells you the truth.

Look at your network segment. Watch your impressions-to-clicks gap. Separate your funnel stages. Do those three things and you'll know what Google's AI rollout is actually doing to your budget, instead of finding out in a quarterly report when the money's already gone.

If you want a second set of eyes on how your search spend is structured heading into the back half of the year, that's a conversation worth having now, while traffic is still warm.