This is one of the better problems an agency can bring to a client meeting, and one of the easiest to communicate badly. The client’s Google organic impressions are down, but the leads coming from ChatGPT are up. Told well, that is a strong story: the metric that matters least fell while the outcome that matters most rose. Told badly, by spinning it, burying the impressions drop, or over-claiming the leads, it quietly erodes the trust that keeps the account. The right approach is honest and confident: lead with the real outcome, explain the channel shift plainly, verify the leads are genuine, and reframe what you report from here. Here is exactly what to say, and what not to.
The short answer
Lead with the leads, because leads are the outcome that matters and impressions are only a top-of-funnel proxy. Explain the shift honestly: AI now answers many informational queries in place, so impressions fall (Pew Research), while high-intent buyers increasingly arrive through AI answers as leads. Do not hide the impressions drop; name it and contextualise it. Verify the leads are genuinely attributable to ChatGPT before celebrating, since attribution is imperfect. Then reframe reporting around outcomes and the new channel, and set the expectation that this continues. Told straight, it is a win to report confidently, not a problem to explain away.
Why this situation happens
Start by understanding the mechanism, because you will need to explain it. AI answers increasingly resolve informational queries directly, so the impressions and clicks a client used to get for those queries decline: Pew found people click a result on just 8% of pages with an AI summary versus 15% without (Pew Research). But the buyers with real intent still need solutions, and they increasingly find them through AI answers that surface and recommend the client, arriving as leads rather than anonymous impressions. So the same shift that cut the impressions is routing the valuable demand through a new channel. Impressions down, leads up is not a contradiction; it is the signature of that shift, and naming the mechanism is the heart of the explanation.
Lead with the outcome that matters
The first communication principle: lead with the leads, not the impressions. Impressions measure how often the client appeared in search; leads measure qualified interest, which is far closer to revenue. So a drop in impressions alongside a rise in leads is a trade of a low-value metric for a high-value one, and your framing should reflect that hierarchy. Opening the conversation with the impressions drop puts the least important number first and frames a win as a loss. Opening with the leads, the real business outcome, frames it correctly. The client cares about customers, not appearances, so lead with the thing that becomes customers.
Do not hide the impressions drop
Leading with leads does not mean hiding the impressions, and this is where credibility is won or lost. Name the impressions drop plainly, because the client can see it themselves, and any attempt to bury or spin it will destroy trust the moment they notice. The honest, strong position is not that nothing declined, it is that what declined matters less than what rose. Stating the drop directly and then explaining it demonstrates that you understand the account and are not managing the client’s perception, which is exactly what earns the benefit of the doubt on the good news. Transparency about the drop is what makes the celebration of the leads believable.
Explain the channel shift plainly
Give the client the why in plain language, not jargon. Something like: search engines now answer many questions directly with AI, so we show up in fewer of the old impression counts for informational queries, but the people ready to act are increasingly finding you through those AI answers and coming to us as leads. This is the same underlying shift behind why is ChatGPT referral traffic dropping and the broader question of is SEO dead because of ChatGPT: the discipline is not failing, the channel is changing. A client who understands the mechanism will trust the story and the metric change that follows, because it makes sense rather than sounding like an excuse.
Verify the leads are real first
Before you celebrate, verify, because over-claiming is the fastest way to lose the trust you just built. Confirm the leads are genuinely attributable to ChatGPT and not miscounted, check how they are being tracked, whether they cite finding the client through AI, and whether the volume and quality are real, since AI attribution is imperfect and some of it hides in direct traffic. Do not assert a clean causal line you cannot support; if the attribution is partial, say so. It is far better to present verified, conservative lead numbers you can defend than impressive ones that collapse under a follow-up question. Honesty about the certainty of the attribution is part of honest reporting.
What to say and what not to say
This table contrasts the credible framing with the trust-eroding one.
| Do say | Do not say |
|---|---|
| Leads are up, here is the verified evidence | Bury the leads under caveats or hype them unverified |
| Impressions fell, here is why | Hide or spin the impressions drop |
| The channel is shifting to AI answers | Everything is fine, ignore the change |
| Here is how we will measure it now | Keep reporting only impressions |
| High-value leads replaced low-value impressions | Impressions are all that matter |
| Attribution is partial where it is | Claim precise causation you cannot support |
Reframe the reporting
The shift demands a reporting change, and you should propose it explicitly. Keep showing impressions and rankings for context, but centre the report on outcomes, leads and conversions, and add the new channel: AI referrals and the client’s presence in AI answers, using an approach like the agency measurement covered in how do SEO agencies measure AEO performance for clients reliably. Explain that fixating on impressions now measures the wrong thing, because impressions for informational queries will keep falling by design as AI answers them. Moving the report’s centre of gravity to outcomes is not moving the goalposts; it is measuring what the client actually cares about, which the old impression-heavy report obscured.
Set expectations going forward
Close the loop by setting expectations, so the next report is not a surprise. Tell the client this shift will continue: informational impressions will keep softening as AI answers more queries, and the work is to capture the high-intent demand that remains, increasingly through AI answers, and to keep converting it into leads. Because ranking still feeds AI answers, with a large share of AI citations coming from top-ranked pages (Ahrefs), the SEO work still matters, it just pays off differently now. Framing the future honestly, more of the same shift, measured by leads, prevents the client from reading each future impressions dip as a failure and keeps the relationship steady.
The one-paragraph version
Here is a script you can adapt. Your Google impressions are down, and I want to be direct about that: AI answers are handling more informational searches, so we appear in fewer of those impression counts. What matters more is that your leads from ChatGPT are up, and I have verified those are real and good quality. In plain terms, we traded low-value impressions for high-value leads, which is the trade we want. Going forward I am going to report on leads and AI presence, not just impressions, because that is what actually reflects your results. Delivered like that, honest about the drop, confident about the leads, clear about the metric change, it lands as competence, not spin.
Handling the skeptical client
Some clients will push back, and that is fine if you are ready. If they distrust the leads, show the verification and the tracking, and offer to tighten attribution further. If they mourn the impressions, revisit the hierarchy: impressions never paid the bills, leads do. If they fear the impressions drop signals decline, explain the mechanism and point to the leads as proof the demand is still being captured, just through a new door, since AI visibility tracks with the authority the client’s SEO built (Ahrefs) and brands are actively surfaced in AI answers (Profound). A skeptical client is not a problem when your story is true and your numbers are verified; it is a chance to demonstrate exactly the rigor that retains accounts.
A worked example
An agency opens a review with a client whose impressions fell 18%. Instead of apologising, they lead: your qualified leads are up 30%, and here is where they came from, ChatGPT increasingly recommending you. Then they name the impressions drop and explain the AI channel shift plainly. They show the lead verification, acknowledging that some AI traffic is hard to attribute precisely and giving the conservative number. Finally they propose a reporting change centred on leads and AI presence. The client, who came in worried about the impressions line, leaves reassured and impressed, because the story was honest, the win was real and verified, and the path forward was clear. That is the difference framing makes.
Common mistakes
The first mistake is leading with the impressions drop, framing a win as a loss. The second is hiding or spinning that drop, which destroys trust when noticed. The third is over-claiming the leads without verifying attribution. The fourth is refusing to change the reporting, so you keep measuring the wrong thing. The fifth is treating the client’s skepticism as an attack rather than a chance to show rigor. Avoid these, lead with verified leads, be honest about impressions, explain the shift, and reframe the metrics, and a scary-looking chart becomes a confident, credible win.
The bottom line
What do you say to a client whose Google organic impressions fell but ChatGPT leads rose? Lead with the leads, because they are the outcome that matters, and be honest about the impressions drop rather than spinning it. Explain the channel shift plainly, AI answers informational queries in place, so impressions fall, while high-intent buyers arrive through AI as leads, and verify those leads are genuinely attributable before celebrating. Then reframe the reporting around outcomes and the new channel, and set the expectation that the shift continues. Told straight, this is a win to report with confidence, and telling it straight is exactly what keeps the client’s trust.
Frequently asked questions
What do I tell a client if organic impressions fell but ChatGPT leads rose?
Lead with the leads, because they are the outcome that matters, while impressions are only a top-of-funnel proxy. Explain the channel shift honestly: AI now answers many informational queries in place, so impressions fall, while high-intent buyers increasingly arrive through AI answers as leads. Do not hide the impressions drop; name it and contextualise it. Verify the leads are genuinely from ChatGPT before celebrating, then reframe reporting around outcomes and the new channel.
Is falling impressions but rising leads a good result?
Usually yes, and often a very good one, because leads are far closer to revenue than impressions. Impressions measure how often you appeared, leads measure qualified interest, so trading low-value impressions for high-value leads is a favourable swap. The caveat is verification: confirm the leads are genuinely attributable to ChatGPT and are good quality before declaring victory, since AI attribution is imperfect and over-claiming damages trust.
Should I hide or downplay the impressions drop with the client?
No. Name it plainly and then explain it. Trying to hide or spin a metric a client can see themselves destroys credibility, while honestly stating the drop and contextualising it as a channel shift builds trust. The strong position is not that nothing fell, it is that the thing that fell matters less than the thing that rose. Lead with honesty, then with the leads, and the impressions drop becomes part of a coherent, credible story.
How do I report this shift going forward?
Reframe the reporting around outcomes and the new channel. Keep showing impressions and rankings for context, but centre the report on leads, conversions, and AI-sourced results, and add tracking for AI referrals and brand presence in AI answers. Set the expectation that informational impressions will keep softening as AI answers more queries, and that the goal is capturing the high-intent demand that remains, measured by leads rather than impressions.
