What to say when your client wants to cut Google Ads because 'everyone asks ChatGPT now'
The email lands on a Tuesday afternoon: "I've been thinking and I reckon we should lower our Google Ads spend. My nephew says nobody searches on Google anymore, that everyone asks ChatGPT now." And you, who has run that client's account for three years and knows full well that Google Ads brings in 60% of their leads, have to decide how to respond without sounding like a salesperson defending their commission.
First things first: the objection isn't absurd. Your client has seen a real change with their own eyes — they use ChatGPT, their kids use it, their company uses it — and they're extrapolating. The mistake isn't in what they see, it's in the scale they assign to it. Your job in that conversation is to separate what's genuinely happening from what's perception, and the data backs you up more than you'd think.
The figure that dismantles the "everyone"
The most solid figure that exists on ChatGPT usage in Spain comes from Funcas, in its III Survey on AI (2026): frequent use of ChatGPT has gone from 4% to 28% in two years.
Read it twice, because there are two arguments inside that figure and your client has only heard one:
- 28% is brutal growth. Multiplying by seven in two years is something almost no channel in the history of digital marketing has managed. On that point your client is right: this is serious and it's moving fast.
- But 72% of Spaniards don't use ChatGPT frequently. Nearly three in four. Classic search is still, by a wide margin, where the majority of their market lives today. "Everyone asks ChatGPT" is, quite simply, false in 2026.
The INE adds context: 37.9% of the population used generative AI in the last quarter of 2025. That's a lot for a three-year-old technology, and it's still a minority. If your client cuts the channel that reaches 100% of searchers to chase the 28% — which, on top of that, doesn't yet have a mature advertising channel — they're doing exactly the opposite of what their own conversion numbers tell them.
There's another, more down-to-earth reason worth saying out loud: today you can't buy ads on ChatGPT the way you buy on Google Ads. The equivalent doesn't exist. Cutting a measurable, profitable channel to "invest in AI" when there isn't yet anywhere to invest that ad spend means ending up with neither.
What is changing (and denying it will cost you credibility)
If your answer is "relax, nothing's happening, Google's the same as ever", you'll lose the next conversation, because something is happening and the data says so too:
- Where Google shows AI Overviews, the average organic CTR falls by 61% in Spain, and traffic arriving from AI grows by 527% in a year (ismajimenez.com).
- Zero-click searches — those that end without a single click — have gone from 56% to 69% in a year (data cited in stucom.com).
- Gartner forecasts 25% less traditional search volume before the end of 2026 (press release, February 2024).
Notice that almost all of that impact lands on organic traffic, not paid: ads still appear and still get clicks. But the underlying trend is real: part of their ideal customer's buying journey — the "what are the options?", "which one do you recommend?" stage — is moving into conversations with AI. And there, their brand either appears or it doesn't, with no bidding budget able to change it.
The honest answer to your client therefore has two parts: "no, it's not time to cut Google Ads yet" and "yes, it is time to start watching what AI says about you."
The script for the meeting
A structure that works, in four steps:
- Validate the instinct. "You're right that this is changing, and I'm glad you raised it: most clients don't see it coming." Starting by granting them a partial point defuses the fight.
- Put the data on the table. Funcas's 28% and its flip side, the 72%. Better still if you bring their own data: how many conversions Google Ads brought them last quarter and at what cost. Against a concrete ROAS, "my nephew says" loses steam on its own.
- Show them what AI says about their business today. This is the move that turns a defensive conversation into a sale. Ask ChatGPT, Gemini and Perplexity what their ideal customer would ask ("best [their service] in [their city]") and bring the screenshots. Most likely they won't appear: in the study we ran on 9,865 Spanish SMEs, 91% appeared in only 1 of the 4 main AIs (full study). That reframes everything: the problem isn't that they spend too much on Google, it's that in the new channel they don't yet exist.
- Propose the complement, not the substitute. Keep Google Ads where it performs and add a new leg: monitoring and working on their AI visibility. Different budgets, different objectives. How to frame that proposal without promising hot air is laid out in what to promise about AI visibility.
The tactical mistake we see repeated: digging in on the "no" and winning the argument this time. The client comes back six months later with the same idea, reinforced by another headline, and that second time they don't ask: they cut. If you channel their concern toward a new service now, you control the narrative; if you block it, they'll eventually resolve it with another agency that tells them what they want to hear. That conversation, once a competitor is already whispering in their ear, is a different one: you'll find it in how to defend the renewal against an agency selling AI.
For step 3 you don't need to set anything up: with a free AI visibility test you'll see in minutes which AIs your client appears in and which they don't, with data you can take to the meeting as is. And if you decide to turn it into a recurring service, Surfeo for agencies repeats that measurement every week for your whole portfolio, so the next time a client raises the topic, the report is already done.
Frequently asked questions
What if my client is right and their specific audience really does use AI a lot?
It can happen: the 28% isn't spread evenly, and among young, technical or very digital audiences the share is higher. But the answer still isn't to cut blindly: it's to measure. Check whether their traffic from AI sources is growing, ask the AIs what their audience would ask, and decide with their data, not with general surveys. If their audience really is there, all the more reason to work on their AI visibility, not just cut Ads.
Can't I just advertise on ChatGPT and solve it with budget?
Today there's no advertising channel on ChatGPT comparable to Google Ads for a Spanish SME. When there is, it'll be one more option; in the meantime, the only way to appear in AI answers is for the sources the AI consults to speak well (and enough) of your client. That's visibility work, not bidding.
How much budget should I shift from paid to AI visibility?
To begin with, none: they're separate lines. Paid stays as long as its ROAS justifies it, and AI visibility comes in as a new service with its own budget, usually a small fraction of media spend. If the client insists it has to come from somewhere, let it come from the worst-performing paid spend, never from the campaigns that sustain the business.
Does this same conversation apply to the client who wants to cancel SEO?
The structure does; the data plays a different role: in SEO the impact of AI already shows up in clicks, and the conversation is more delicate. The version for that case is in what to say to a client who wants to cancel SEO over ChatGPT.
Before your next meeting with that client, spend five minutes looking at the figure that's going to change the conversation: run the free AI visibility test on their site and bring the result printed.
