AI chatbots intercepting Google Ads leads before the click

AI assistants like ChatGPT, Claude, Gemini, and Perplexity are answering buyer questions before they ever become Google searches — and it’s changing what a Google Ads lead costs and what it’s worth. Here’s what the data actually shows, and how to fix it.

Your cost per lead is up. The forms keep coming in — but the people behind them don’t answer the phone, don’t live in your service area, or were never real prospects to begin with. If that sounds familiar, you’re not imagining it, and you’re not alone.

There’s a reason this is happening now. Millions of buying decisions that used to start with a Google search now start with a question to ChatGPT, Gemini, Claude, or Perplexity — and every one of those conversations is a search your ads never had a chance to win. What’s left in the auction is more expensive, harder to qualify, and easier for automated bidding to get wrong.

The Short Answer

Google Ads isn’t dying — but your lead quality might be. AI now answers the research questions that used to become clicks, cost per lead keeps climbing, and if the only signal you send Google is a form fill, automated bidding will happily buy you more form fills from people who will never become customers. The fix is better inputs, not a bigger budget.

The data suggests something important is changing — but the simplest explanation, “people stopped using Google,” is not supported. Google Search remains enormous, and Alphabet continues to report strong growth in Search advertising. The more accurate story is that consumer research is spreading across traditional search, Google’s own AI answers, and standalone chatbots. That shift can remove or delay clicks, change the queries that reach advertisers, and expose weaknesses in conversion tracking that previously went unnoticed.

Is Google Ads Collapsing — or Just Getting More Expensive?

Alphabet’s financial results do not resemble a platform in free fall. Google Search and Other advertising revenue grew 12% year over year in Q2 2025, 15% in Q3, and 17% in Q4. In 2026, Google said Search queries had reached an all-time high as AI Mode expanded.

That top-line strength matters because it prevents an exaggerated conclusion: AI chatbots have not replaced Google Search at scale.

But advertisers don’t experience the market through Alphabet’s total revenue — they experience it at the campaign and sales-pipeline level, which is exactly where we live every day. LocaliQ’s analysis of more than 16,000 U.S. search campaigns found the average cost per lead rose from $66.69 in 2024 to $70.11 in 2025 — a 5.13% increase — with cost per click rising in 87% of industries studied. In home services, 69% of businesses saw cost per lead increase, with the average rising 10.51% year over year.

Here’s what those benchmarks don’t tell you — and what we flag in every client audit we run: they measure recorded leads, not whether the prospects were qualified, reachable, or profitable. A campaign can show an acceptable conversion rate while the business receives more spam, job seekers, duplicate inquiries, price shoppers, or consumers outside its service area.

How Is AI Absorbing the Research Journey Before the Click?

Independent browsing data shows why the funnel feels different. In a March 2025 study, Pew Research Center found that Google users clicked a traditional search result in just 8% of visits when an AI summary appeared, compared with 15% when no AI summary appeared. Users clicked a source inside the AI summary in only 1% of visits.

That study focused on outbound links generally, not paid-ad clicks specifically. It nevertheless establishes an important behavioral mechanism: when an answer is delivered directly on the results page, fewer people continue to another website.

Standalone chatbots create a second form of displacement. A consumer can compare products, understand terminology, build a shortlist, or decide what questions to ask — all without generating a single Google ad impression. The user may later return to Google with a narrower, more commercial query, or proceed directly to a company already mentioned by the assistant. Either outcome reduces the value of owning only the traditional top-of-funnel keyword.

This does not mean every chatbot session replaces a Google search. AI assistants are also used for writing, coding, summarization, and personal tasks. Search engines still attract far more visits, and different datasets count prompts, searches, visits, and users in incompatible ways. Market-share figures should be read as directional evidence, not interchangeable measurements.

Why Does Gemini’s Growth Matter So Much?

Gemini is not merely another competitor taking traffic away from Google. It is Google’s strategy for preserving its role as search behavior changes — and the numbers show how fast that strategy is working.

Google reported more than 400 million monthly active Gemini app users in May 2025, more than 450 million in July, more than 650 million in October, more than 750 million by February 2026, and more than 900 million by May 2026. That is more than a doubling in one year. Google also reported that AI Overviews had more than 2.5 billion monthly users and AI Mode had surpassed one billion monthly users by May 2026.

Similarweb’s global web-traffic panel shows the competitive effect. Between June 2025 and May 2026, ChatGPT’s share of visits within the generative-AI website category fell from approximately 76% to about 53%, while Gemini rose from below 9% to roughly 27–28%. Claude increased from around 2% to nearly 9%. These figures cover website traffic within the AI category; they do not include every app interaction and should not be confused with total search-market share.

ChatGPT remains a massive destination. OpenAI reported 700 million weekly active users in September 2025, more than double the 400 million reported that February. The important point for advertisers is not which assistant “wins.” It is that conversational discovery now occurs across several rapidly growing platforms — which is why our SEO and GEO strategies now target visibility in AI answers, not just blue links.

Why Can Lead Quality Fall While Google’s Numbers Rise?

In the campaigns we audit and manage, we consistently see four forces working together:

AI answers satisfy more early-stage intent. The remaining paid-search audience may contain a different mix of urgent buyers, repeat comparison shoppers, and ambiguous queries.
Google is integrating AI rather than surrendering the journey. AI Overviews and AI Mode keep users inside Google’s ecosystem, and Google is adding ads to those experiences. This protects Google’s advertising opportunity but changes placement and context.
Automation optimizes toward the easiest measurable action. Google itself says Performance Max is only as good as the inputs it receives. If a basic form is the primary conversion, bidding learns to find form submitters — even when many are poor prospects.
Rising competition increases cost without improving intent. More bidders pursuing a limited pool of commercially valuable clicks raises both cost per click and cost per qualified opportunity.

For this reason, AI disruption and low-quality leads should not be treated as a proven one-to-one causal relationship. The same symptoms can come from broad match, Performance Max expansion, weak negative-keyword controls, geographic leakage, fraud, landing-page friction, call handling, price changes, competitors, or inaccurate conversion settings. Diagnosing which of these is actually hurting a campaign is precisely the kind of forensic account work we do before recommending a single change. AI is one structural contributor — not an all-purpose diagnosis.

What Metric Should Replace Cost Per Lead?

Advertisers should separate three stages that are too often combined:

Platform conversion: a call, form, message, or other action counted by Google Ads.
Qualified lead: a real prospect in the correct market who needs the service and can be contacted.
Acquired customer: a qualified prospect who produces revenue and acceptable margin.

If Google receives only the first signal, its automation cannot reliably optimize for the third. The solution — and the foundation of how we structure conversion tracking for our clients — is to send offline outcomes (qualified, quoted, booked, sold, and revenue) back into the advertising system using enhanced conversions for leads, CRM imports, or call-tracking integrations. Campaigns should be judged by cost per qualified lead, close rate, customer-acquisition cost, and revenue — not by form volume alone.

What Should Advertisers Do Right Now?

Google Ads should remain part of the plan where it produces profitable customers — especially for local services and urgent, high-intent needs. But it should no longer carry the entire burden of discovery. Here is the playbook we implement for the businesses we work with:

Tighten paid-search control. Audit search terms, location settings, network expansion, negative keywords, conversion actions, and Performance Max traffic.
Feed Google better outcomes. Import qualified-lead and closed-sale data so bidding is rewarded for business value rather than superficial activity.
Build AI-readable authority. Publish direct answers, original expertise, case studies, service-area information, strong entity signals, and accurate structured data — the same schema and content architecture we build into every site we design.
Measure AI discovery separately. Track referrals and assisted conversions from ChatGPT, Gemini, Perplexity, Claude, and other identifiable sources, while recognizing some influence will remain unattributed.
Protect brand demand. Maintain consistent facts across the website, business profiles, trusted directories, and third-party coverage.
Diversify acquisition. Use email, remarketing, social, referral programs, and content so every new customer does not depend on the same paid auction.

The Bottom Line: What Standard Should Google Ads Meet Now?

AI is not destroying Google Ads overnight. It is changing what happens before the click, which searches still produce a click, and how much reliable intent costs.

Google remains dominant and is using Gemini to transform Search from within. ChatGPT remains the largest standalone chatbot by many usage measures, while Gemini and Claude are gaining share within the AI category. For advertisers, the market is no longer a simple contest between Google and one rival. It is a fragmented discovery environment in which consumers can obtain answers without visiting a website — and platforms can optimize campaigns toward conversions that never become customers.

The companies that adapt will stop asking only, “How many Google Ads leads did we receive?” They will ask, “Which channels influenced this customer, was the lead genuinely qualified, and did the investment produce profitable revenue?”

That is the standard Google Ads must now meet — and it’s the standard we hold every campaign to.

Are Your Google Ads Producing Customers — or Just Conversions?

For over 25 years, AdverGroup has helped Chicago-area businesses connect paid search, SEO, GEO, structured data, conversion tracking, and content strategy — so marketing is measured by customers, not just clicks. Let us audit your campaign and show you where qualified leads are leaking.

Request a Campaign Audit or call (847) 401-6733