For most of the last decade, an agency's job was straightforward: rank the client on Google, run their ads, manage their content. The conversation was about keywords, backlinks, and CTR.

That conversation is shifting fast. Over 700 million people use ChatGPT every week. Perplexity now answers 100+ million queries a day. AI-driven referral traffic to websites grew 527% year-over-year. And buyers have started asking AI tools the questions they used to type into Google: "best CRM for a 20-person SaaS company", "top dermatologist in Milan", "what's the most reliable solar panel installer in Bristol?".

If the client doesn't show up in those answers, the agency is going to hear about it.

This piece is for agency owners and account leads who want to be ready for the conversation before it lands in their inbox. Here are the seven questions your clients will start asking in 2026 — what each one really means, and how to answer it without bluffing.

1. "Why doesn't ChatGPT mention us?"

What they're really asking: Am I invisible to the next generation of buyers, and is that your fault?

This is the question that opens the door to every other one. The honest answer almost always has the same shape: ChatGPT, Perplexity, Claude and Gemini work nothing like Google. They don't crawl your site the same way. They don't weight your backlinks the same way. They don't read your meta description.

The technical fix list is usually a mix of: AI crawlers blocked by robots.txt or Cloudflare Bot Fight Mode, missing or stale schema markup, no llms.txt file, citations from low-authority sources, and content structured for Google's featured snippets rather than for an LLM's answer synthesis. We've covered the most common technical traps in 8 AI Invisibility Traps That Stop ChatGPT From Recommending Your Business.

The strategic answer for the agency is harder: "We've been optimizing for Google. AI search is a different surface and needs its own work. Here's what we're going to do about it."

2. "Should I be worried about losing organic traffic?"

What they're really asking: Is my Google strategy still working, or do I need to panic?

The short version: Google AI Overviews (the "AI snapshot" at the top of the search results page) is already cannibalizing clicks for informational queries. For purely transactional queries, organic is still healthy. For research and comparison queries — the ones that drive the top of your client's funnel — AI surfaces are eating share fast.

Don't tell the client to panic. Tell them the funnel is splitting in two: a Google-driven transactional layer and an AI-driven research layer. Both need to be optimized, but the playbook is different. The clients who win in 2027 will be the ones who started in 2026.

3. "Can you measure my AI visibility the way you measure my SEO?"

What they're really asking: Can you give me a number, a dashboard, and a monthly report so I know if we're winning?

Yes — and this is the question that's hardest to bluff. You either have the tooling to answer this or you don't.

What the client needs to see, monthly: a visibility score across the major AI platforms (ChatGPT, Perplexity, Claude, Gemini, Mistral, You.com), a share-of-voice comparison against named competitors, a sentiment trend, and a list of the specific queries where the client appears or doesn't. If you can't produce that, you're going to lose the account to an agency that can.

This is increasingly table-stakes. The market has already moved: tools that measure AI visibility for agencies exist at every price point, from $29/month to $50k/year enterprise contracts. The competitive question for your agency isn't whether to offer AI visibility reporting — it's how soon, and on whose tooling.

4. "What about Bing Copilot? And Grok?"

What they're really asking: Are you keeping up with every new platform, or are you going to be caught flat-footed in six months?

The honest answer is that ChatGPT, Perplexity, Claude, and Gemini account for the overwhelming majority of branded AI search volume in 2026. Bing Copilot has nontrivial enterprise reach via Microsoft 365. Grok is small but growing inside the X ecosystem.

Your real positioning here is to be ahead of the platform conversation, not behind it. Build the answer that says: "We optimize for the platforms that account for 90%+ of AI search today, and we monitor the emerging ones. When one of them crosses a usage threshold worth your attention, we'll bring you the data and the plan."

That's a defensible position. "We optimize for everything" is not.

5. "Can you guarantee we'll rank in ChatGPT?"

What they're really asking: Is this another channel I can pay you to control, or is it a gamble?

This is the question where agencies that overcommit will get burned. There is no "rank #1 in ChatGPT." ChatGPT generates an answer per query, on demand, partly from training data, partly from web search, partly from internal heuristics that change month over month. You can't buy placement. You can't game it with link farms. You can dramatically improve the odds of being included in the model's answer set — but you can't promise it.

The answer that works: "We can't guarantee placement, and you should be skeptical of anyone who says they can. What we can do is measure where you stand today, identify the technical and content gaps that are keeping you out of answers, fix them, and track the lift over time. That's the same shape as SEO — you don't pay Google for rank either."

6. "What about the smaller AI tools my customers use?"

What they're really asking: Are you thinking about my actual customer's behavior, or just the headline AI tools?

The big names dominate volume, but vertical AI assistants (legal, medical, B2B SaaS specific) are starting to matter for niche industries. A B2B SaaS client whose buyers use specialist research tools may need a different visibility strategy than a consumer brand whose buyers use ChatGPT.

The right move is to start with the major platforms (where 90%+ of branded AI search lives in 2026), then layer in vertical assistants as the client's specific buyer behavior calls for it. Don't pretend you can optimize for every long-tail AI surface. Do show the client you know their buyer's actual research path.

7. "What's this going to cost?"

What they're really asking: Is this a new line item I have to find budget for, or is it included?

This is the agency positioning question, not a pricing question. The agencies that handle it well in 2026 will fall into one of two camps:

Camp A: AI visibility is bundled into the existing retainer. The agency absorbs the tooling cost, treats AI visibility as part of organic strategy, and reports on it as a single channel. Easy for the client. Tight margins for the agency unless tooling is efficient.

Camp B: AI visibility is a separately priced service line. The agency invests in deeper tooling, builds out AI-specific deliverables (audit reports, citation tracking, sentiment dashboards), and charges $500-$2,000/month per brand as a discrete add-on.

Both models work. The mistake is camp C — offering AI visibility because clients are asking without a real service definition or pricing structure behind it. That's where margins evaporate and the conversation goes sideways.

What this means for your agency in 2026

The agencies who win this year aren't going to be the ones with the loudest "we do AI SEO" marketing. They're going to be the ones who can sit across the table from a client, hear the question "why doesn't ChatGPT mention us?", and answer it with: data, a fix list, a monthly report, and a price.

If you want to see what that looks like end-to-end — what an agency-tier audit covers, how the white-label reporting works, what the agency dashboard looks like — see VisibAI for Agencies.

If you're earlier in the journey and want to understand the basics first, What is AI Visibility? is the right starting point.

The clients are going to start asking. Be ready.