How AI Overviews and AI Mode are changing Google Ads in 2026

If your Google Ads account feels different this year, you're not imagining it. Google now answers a growing share of searches with an AI Overview right at the top of the page, and AI Mode turns whole sessions into conversations instead of ten blue links. For advertisers, the question isn't whether this matters. It's what to do about it.

I run client accounts through this shift every week, and here's my honest summary of what's changed, what hasn't, and where I'm putting budgets now.

What actually happened to the search results page

Two years ago, most searches ended with a list of organic results and maybe some ads. Now, a large chunk of informational queries show an AI-generated answer first, with sources cited underneath. Google's AI Mode goes further: follow-up questions, comparisons, and recommendations generated on the fly, similar to chatting with an assistant that happens to search the web.

For searches with clear buying intent, ads still show, and they often look exactly like before. That part matters. The change is heaviest on research-style queries: "what is", "how does", "best way to", comparisons, and early-stage questions. Those results increasingly get consumed by the AI answer box.

So the first mental shift is this: the top of the funnel is being absorbed by AI answers, while the bottom of the funnel still behaves a lot like search always did.

The metric that stopped meaning what it used to

The biggest reporting trap right now is impressions. Accounts across the industry have seen impression volumes wobble on informational keywords, not because demand disappeared, but because the answer now satisfies people without a click, or shows in formats that count differently. Impressions were always a vanity metric; they're even more misleading when the layout itself is shifting underneath you.

What I watch instead:

• Conversions and cost per conversion. Boring advice, but it cuts through all of this noise. If leads keep coming at an acceptable cost, the sky is not falling.
• Click-through rate on commercial keywords versus research keywords. Split them into separate campaigns if you haven't. The trends look completely different.
• Search terms report quality. Longer, more conversational queries are showing up. Some convert surprisingly well because they describe the problem in detail.

What I'm changing in real accounts

Here's the practical playbook, based on what's working for my clients rather than theory:

1. Pull budget away from pure research keywords. Generic "what is X" terms cost money and increasingly feed the AI answer instead of your landing page. High-intent terms with buying language behind them deserve that budget.
2. Feed the machine better inputs. AI answers cite and summarize content. Landing pages with genuinely useful information, clear pricing signals, and real expertise stand a better chance of being referenced, and they convert better regardless. This rewards businesses that took content seriously and punishes thin pages.
3. Keep conversion tracking boring and bulletproof. As layouts shift, your tracking is the one stable source of truth. Enhanced conversions and offline import matter more now, because some conversions arrive later and through stranger paths than they used to.
4. Treat Performance Max with open eyes. PMax follows attention across surfaces, including AI placements. It can work well, but only with strong assets and clean conversion data. Weak inputs produce weak output faster than ever.
5. Broaden match types carefully on proven themes. Conversational queries are longer and more varied, which favors broader matching on themes you know convert, paired with tight negative keyword lists to keep the junk out.

The opportunity hiding inside the panic

Every platform shift shakes out the lazy competitors. A lot of small businesses read a scary headline about AI search, assume ads are dead, and turn everything off. Their competitors who stayed calm picked up cheaper auctions and steadier lead flow. I've watched this movie before with mobile, with local packs, with every big layout change. The advertisers who measure instead of panic end up owning the channel.

There's also a genuine opening in being cited. When an AI answer references a source, that mention carries trust a plain ad doesn't. Businesses with real content, real reviews, and clear expertise are getting pulled into those answers. That's earned media sitting inside a paid channel, and most of your competitors haven't noticed yet.

What I would NOT do right now

• Don't slash budgets blindly. Check cost per conversion first. Most of my accounts are paying similar or lower costs per lead than last year, even where impressions moved around.
• Don't buy every "AI SEO" service that promises citations in overviews. Nobody outside Google controls those answers reliably. Good content and good products are still the input that matters.
• Don't rebuild your account structure every month chasing the news. Stable structure plus honest measurement beats constant churn. Make changes you can attribute.

The bottom line

Search is splitting into two jobs: answering questions and facilitating purchases. AI Overviews and AI Mode own more of the first job every quarter. Paid ads still own much of the second, especially for local and service businesses where the searcher wants a provider, not a paragraph.

Watch costs per lead, protect your tracking, feed the system honest data, and put your budget behind keywords where someone actually wants to buy. Do that, and these changes become a filter that removes your weakest competitors instead of a threat to yours.

If you want a second pair of eyes on how AI search has shifted your specific account, that's exactly what my free strategy call is for. I'll tell you plainly whether anything needs to change or whether you're already positioned well.