How to Sell GEO Services: A Practical Guide for Agencies

At some point this year (definitely before the World Cup, which is how I measure the passage of time), the tone of the generative engine optimization (GEO) conversation between client and agency began to shift. 

Like the sluggish movement of tectonic plates, it took a while for the most common question to move from ‘what is GEO?’ to ‘OK, but what should we do, and why?’ But if you live and work in Agency Land, that shift can feel as seismic as South America breaking away from Africa, or Australia drifting off into the distance to do its own thing.

I’ll spare you any more metaphors for now. The point is that your clients have woken up. Not just to the fact that their customers are asking LLMs about their products, services and industries, but to the fact that they need somebody to help them show up there, for agents and audiences alike.

What’s still up in the air is who they’ll turn to first, and why they should trust that person’s judgment when they do.

I was deliberately vague earlier when I said ‘Agency Land’. Whether you’re a PR agency, a web dev shop, a content team or an SEO team, that somebody could be you. Cara Corbett (our VP of Partner Growth) and I spent our last Partner Club webinar outlining exactly how you make that happen: what you’re actually selling, when to raise it, the questions to ask, the proposal that closes it, and what to say when clients push back. 

In this article, I cover all of these topics in detail, so you know exactly how to sell GEO services to your own clients.

Why GEO Is a Service You Can Sell Today

If your clients have woken up, it’s because their customers got there first, and the numbers show just how far ahead they are.

Half of consumers already use AI-powered search, according to McKinsey. About 60% of searches now end without a click, per Bain & Company. People get their answer on the search results page and never visit a website. The market is nearly doubling. The Business Research Company values the SEO services market at $108 billion this year, rising to $203 billion by 2030. The stat I keep coming back to, though, is about ownership rather than demand.

• 50% of consumers already use AI-powered search
• ~60% of searches now end without a click
• $203B SEO services market by 2030, nearly double this year’s $108B

Muck Rack’s 2026 State of PR report found that nearly three-quarters of PR professionals see GEO as important to their strategy, yet 29% say no one in their organization owns it, and another 13% aren’t sure who does. And before anyone in content, brand or SEO feels smug about that, we see exactly the same pattern on your side of Agency Land too.

• 29% of PR teams say no one in their organization owns GEO

“The demand for GEO is there, and so is the budget. What’s missing is someone in the organization who takes ownership of it, and that’s the opening for agencies.”

Cara Corbett, VP of Partner Growth, SUSO Digital

So the demand is real, the budget exists, and nobody has planted a flag yet. You already have the client relationships and many of the skills that feed AI visibility. The agencies that win the next few years will be the ones who can walk into a meeting and say:

1. Here’s your current performance.
2. Here’s the gap between you and your competitors.
3. Here’s what’s causing that gap.
4. Here’s what we propose to fix it.

If your GEO sales pitch already answers ‘how do we show up in ChatGPT against our competitors?’, you’re having a completely different conversation from the agency whose deck doesn’t. That’s as true at a renewal as it is in a new business pitch.

What You’re Actually Selling: The Four Pillars of GEO

Here’s the good news: you don’t need to be a GEO expert to sell this. If you know how to sell SEO services, you’re already most of the way there. Here’s the slightly less good news: you do need to know what’s inside the box, otherwise the client’s very first follow-up question will catch you out (and that’s sales 101, which I suspect nobody reading this needs me to explain). To keep it simple, we break GEO into four pillars:

• Prompt monitoring: how you measure whether any of this is working.
• Earned and PR: where your client gets mentioned and cited elsewhere on the internet.
• On-site content: what’s on the client’s own site, and whether an LLM can actually use it.
• Technical: whether AI bots can get into the site in the first place.

Most of Agency Land already covers one or two of these, which makes them the natural place to start the conversation. The rest can be picked up by a white label partner, so the client still gets the full program and you get to lead with what you do best.

1. Prompt Monitoring

Prompt monitoring always comes first, because everything else gets judged against it. It tracks the prompts people put into ChatGPT, Gemini, Claude, Perplexity and the rest in your client’s space, and shows whether your client turns up in the answers, and if they don’t, exactly who’s turning up instead.

The catch is that nobody can see the exact prompts people type. So you track a representative set, and if you pick the wrong ones, the whole dataset skews. Plenty of people use AI to generate their prompt sets, which is completely understandable, but it can very quickly turn into a snake eating its own tail: a closed loop of prompts that no human being has ever actually asked.

We build ours from customer personas, real search data and the client’s own first-party data instead: what their customers actually ask in sales calls, support tickets, reviews and on-site search. For a New York moving company, we split ‘best movers NYC’ by intent, neighborhood and situation, and the prompt set that came out of it didn’t just tell us what people were looking for. It handed us a content plan.

Monitoring is also what turns GEO from a one-off audit into a proper retainer, because it never really stops. And because it tracks competitors, it produces the single most persuasive thing you’ll ever put in front of a client: a rival being named in an answer where they should be.

So get the prompt set right before you report a single number – everything else in the program stands on it.

2. Earned and PR

PR agencies, you’ve had a head start here, whether or not anyone told you. LLMs train on and cite exactly the kinds of sources you already place clients in: news coverage, industry publications and authoritative third-party sites. What GEO adds is a way to measure that value, plus a few tactics that make the same coverage work harder in AI answers. (We’ve written a full AI search guide for PR agencies if you want to go deeper.)

The thing I’d really stress is framing. If your client hears ‘the same PR strategy with AI bolted on’, you won’t get the upsell. So be clear about what’s genuinely different:

• Targeting. You prioritize the sources LLMs actually cite. That often goes beyond a normal media list: Reddit (which surprises most people), Wikipedia, Wikidata, LinkedIn and industry review sites, depending on the prompt set, vertical and goals.
• How it’s written. Coverage needs to be easy for an LLM to extract and cite, and it should link back to the client’s site so the two reinforce each other.
• Nothing is lost. Every placement still works as a backlink for SEO as well as a mention for AI.

When a global crypto exchange relaunched in the UK, we optimized their existing outreach strategy for GEO. That same PR work went on to pick up around 4,400 AI mentions.

In other words, it’s the work you’re already doing, measured properly and pointed at the sources AI actually reads.

3. On-Site Content

As a former head of content, I’ll admit this is the pillar closest to my heart, so I’ll try to keep it brief (no promises). LLMs exist to answer questions, so the content that gets cited is the content that clearly answers what people are asking, along with the query fan-outs AI tools generate behind the scenes.

Query fan-out, very quickly: if the central question is ‘best business hotels in London’, a site that also covers ‘best business hotels in London with free breakfast’ or ‘with hot-desking space’ gives the topic more body, and more authority in the eyes of the models.

The basics:

• Clear sections, with key takeaways up top
• The main answer within the first 100-250 words
• Proper FAQs
• The right title tags and H1s

Then there’s the authority layer: topical depth, regular fresh publishing, real named authors and – the one I push hardest whenever I’m talking to clients – corroboration. Other sources backing up what you’re saying. Hardly anyone does it, mostly because traditional content strategy never asked them to. (I’ve covered the rest of the list in Writing for Robots, for anyone who wants the long version.)

Content agencies, most of this comes down to the brief. Plenty of briefs read beautifully and ask for none of the above. For one healthcare brand, we added the medical team’s credentials, proper author bios and peer-reviewed sources, and fixed some crawl issues alongside it. AI traffic rose nearly 600%, and AI conversions more than 300%.

If you’re already producing good content, you’re closer than you think. The question is whether your brief is asking for the right things.

4. Technical

Technical tends to be the scariest pillar in traditional Agency Land. I can only speak for myself there, which is exactly why SUSO has a whole team of people far smarter than me looking after it.

Web and dev agencies, first of all, I’m in awe of you. Secondly, this is where you’ve got a real advantage, because you’re already in the CMS and the codebase. The areas that matter:

• Crawlability. Can AI bots actually reach the content? Plenty of sites still carry robots.txt rules written years ago to block scrapers, which now block the crawlers behind ChatGPT, Claude and the rest. Age gates (on alcohol brands, for example) can stop AI citing content altogether.
• Structured data. Schema tells an LLM who the brand is, what it sells, who’s behind it and what it covers.
• Entity association. The brand needs to be clearly linked in the data to the topics it wants to own.

PR and content agencies without technical people in-house: honestly, don’t let this stop you selling GEO. A white label partner can handle it for you.

Just check it early. Clients can’t see technical issues, but those issues decide whether they show up at all.

When to Raise GEO: Four Trigger Moments

There’s a version of the GEO conversation that feels like a natural next step, and a version that feels like a cold upsell nobody asked for. The difference is almost always timing. These are the four moments when the door is already half open:

• The reporting call. You’re already in the data with the client. A single AI visibility slide is the easiest door-opener you have.
• When the client asks about AI search or ChatGPT. They’ve done half the work for you. Take it away, run a quick audit and come back with insight.
• Renewals and scope reviews. Add a new dimension instead of renewing the same thing, and position GEO as additive to what you already do.
• Campaign and product launches. AI is another discovery surface, and the four pillars can make a launch go a lot further.

“If you’re already walking a client through their monthly performance report, adding one AI visibility slide is the easiest way to open the GEO conversation.”

Cara Corbett, VP of Partner Growth, SUSO Digital

If you only do one of these, make it the reporting call. One slide costs you nothing, and it lets the data start the conversation for you.

The Seven Discovery Questions You Shouldn’t Skip

Discovery is the bit of the process everyone knows matters and nearly everyone rushes, especially when you’re the one raising GEO and the client hasn’t set aside time for it. But even a couple of these answers changes everything that comes after. Cara calls these the non-negotiables, and I’m not going to argue with her:

1. What are your goals for GEO, and which KPIs do you measure? Some clients care about reputation: how AI describes them and whether it gets the facts right. Others want leads and sales. Those goals need different strategies, prompt sets and reporting, and those GEO KPIs become the criteria you’ll be judged on.
2. Which products, services or markets matter most? Markets are the one people forget. Prompts get expensive as scope grows, especially for big brands with lots of segments, so pin it down early.
3. What’s a new customer worth to you? The most useful selling question, and the one agencies skip most. Once you know the answer, you can show how many new customers AI search needs to bring in for the program to pay for itself, and the retainer stops looking like a cost.
4. Who’s your target audience? AI search is conversational and context-driven. The more specific you are about the customer, the more accurate your prompts and your output.
5. Who are your competitors? This sets the benchmarks that matter to the prospect. The competitors they name often aren’t the ones showing up in LLM answers, which is a powerful insight in itself.
6. What’s your timeline? A launch, campaign or internal deadline creates urgency. It also helps you set realistic expectations, because GEO builds over time.
7. What’s your budget? Awkward, but it saves everyone time and lets you shape the right package, whether that’s a focused pilot or a full program.

If you only nail two, make them the first two. They decide which prompts you monitor, and the prompts decide the focus of the whole campaign.

And before you hang up, book the follow-up. Tell the prospect your team will run some preliminary analysis, then lock in the findings call there and then, with the decision makers in the room. It saves you weeks of chasing over email, and it means the people holding the budget hear the opportunity firsthand rather than as a secondhand summary.

“Book the findings call before you hang up on the discovery call. It keeps the momentum going, and you’re not stuck chasing the client over email for weeks.”

Cara Corbett, VP of Partner Growth, SUSO Digital

Selling With Insight: Lead With the Gap

Here’s something I’ve learned the hard way: executive teams don’t really care how query fan-out works. They care whether they’re showing up, whether AI is describing them accurately, and whether they’re beating the people they least want to lose to. So once you know what the client wants and who they’re up against, lead with the competitive gap, not the four pillars.

The fastest way to find that gap is SUSO’s free AI Search Visibility Checker. It assesses how a brand shows up across ChatGPT, Gemini, Perplexity and Claude using 100+ signals, and gives it a score out of 100. There’s no login, and the report lands in your inbox in about five minutes.

It looks at six areas:

• Answer frequency: how often AI tools mention the brand when someone asks about the industry.
• Competitor comparison: the most useful section for opening a sales conversation, because it shows where your client is being left behind.
• Online reputation: whether AI describes the brand positively, negatively or neutrally.
• Brand authority: how well known and well understood the brand is across the web.
• Training data inclusion: whether the brand appears in the sources models learn from, like Reddit.
• Technical accessibility: whether bots can find what’s on the site. (Web agencies, this one’s yours.)

Here’s what that looks like in practice. We ran the checker on a multinational shoe brand, and it showed up in just 1% of the prompts we tested. The report showed which competitors were dominating the conversation, but it also surfaced something far more basic: the brand’s robots.txt was blocking AI bots, so its content and products couldn’t be cited at all. Five minutes to find, a huge insight for the client, and a fast win once it was fixed – and a PR agency looking only at coverage would almost certainly never have spotted it.

For a crypto trading platform, red flags raised by a report just like this one (followed by a deeper analysis) led to work that doubled their AI traffic in six months and added 151 AI citations.

And if you need more depth than a five-minute check can give you, our team can produce a custom report of up to 30 pages, white labeled in your branding if you prefer.

The moment a client can see where they’re invisible, and why, you don’t have to explain why GEO matters anymore. They can see it with their own eyes.

The Winning GEO Proposal Structure

Most GEO conversations die somewhere between ‘that’s interesting’ and ‘send me a proposal’. That gap is where the proposal itself has to do the heavy lifting.

Every agency has its own way of building one, and I’m not going to pretend there’s only one right answer. But this is the structure Cara has used to sell GEO and SEO retainers throughout her career, most recently closing one in under a week, from first call to signature. Take what’s useful:

1. Credibility. Reintroduce yourself (there may be new decision makers in the room) and show why you’re the one to trust: track record, awards, relevant client logos and wins. For GEO, we highlight our talks at LA and New York Tech Week and our webinars with names like Profound.
2. Their pain points and goals. Restate what they told you in discovery. It proves you understand their business, and the recommendation can then hinge on solving those problems.
3. Key findings. This is where the checker or audit output lives: the competitor gap, the technical issues, the ‘here’s why you’re invisible right now’ moment. Lead with their strongest metric, show the tension and frame it as an opportunity, not a failure.
4. What you’re proposing. Not a generic monthly service list. Walk through exactly what they’ll get, tied to the gaps you just showed, with the four pillars turned into real deliverables.
5. Proof. Case studies, ideally from the same industry or a similar problem. Partner Club members can get these from us at any time, fully white labeled.
6. Next steps. What happens between signing and kickoff, what onboarding looks like, and how often you’ll report. Confirm when they plan to make a decision.

“This six-part proposal structure closed a GEO retainer for me in under a week, from the first discovery call to signature.”

Cara Corbett, VP of Partner Growth, SUSO Digital

If every section points back to the gap you found in discovery, the proposal stops reading like a menu and starts reading like a plan.

How to Handle the Four Most Common GEO Objections

Even with a great deck, you’ll get pushback. Objections are everyone’s favorite part of life in Agency Land, I know. These are the four we’ve heard most over the last couple of years, and how we answer them.

1. ‘Isn’t This Just SEO?’

We still hear this one constantly, and as a search agency, I understand why it comes our way. My advice is to start by partially agreeing. If you try to argue that the two are completely different in every way, you’ll lose the room. Good content, topical authority, E-E-A-T and a fast, crawlable site matter for both, and original research earns links for SEO and citations in AI answers alike.

Where they differ is what they’re optimizing for. SEO is about ranking in results. GEO is about being cited in the answer. (We’ve broken down the full differences in SEO vs GEO: What’s the Difference, and Does It Matter for Your Strategy?) And that difference looks a bit different depending on which corner of Agency Land you’re standing in:

Agency type	SEO focus	GEO focus
Content	Keyword research, content gaps	Prompt research, extractable structure, named authors
Web/dev	Crawl budget, Core Web Vitals	AI bot access (GPTBot, ClaudeBot), rendering checks
PR	Backlinks, domain authority	Mentions on Wikipedia, Reddit, review sites

The gap between them is getting wider, too. Ahrefs found that in July 2025, around 76% of pages cited in Google’s AI Overviews also ranked in the traditional top ten. By early 2026, that was down to 38%. Ahrefs puts that mainly down to Gemini 3 and query fan-out, but the direction of travel is hard to argue with. Page one no longer guarantees you a place in the answer, and we’ve seen plenty of brands with brilliant rankings and very little AI visibility.

One extra nugget, because it took a ridiculous amount of research to find. If you want to predict what Claude cites, you’re better off looking at Brave Search than Google: the overlap between Claude’s top-cited sources and Brave’s top ten is roughly 80%, according to Profound. (Different models, different sandboxes – we go into that in our piece on multi-LLM optimization.)

The way I’d put it to a client: ‘If we’re already doing SEO together, this isn’t starting again. It’s taking that work into AI search. A lot of it overlaps, but it’s measured differently, so we can’t assume one equals the other.’

2. ‘Is There Actually Any Traffic? Can You Prove ROI?’

We hear this one on an almost daily basis, and it’s a completely fair question. If it were my money, I’d want to know too. The good news is that you can answer far more of it today than you could even six months ago.

Start with where the attention has gone. Around half of Google searches now show an AI summary, according to McKinsey (anyone who’s used the internet recently will have noticed). Ahrefs’ study of roughly 300,000 keywords found the top result loses around 58% of its clicks when an AI Overview appears, and Pew found users end their search 26% of the time after seeing one, against 16% without. The demand hasn’t disappeared. It’s moved into the answer, which is exactly where GEO puts your client.

• ~50% of Google searches now show an AI summary
• ~58% fewer clicks for the top result when an AI Overview appears
• 26% vs 16% of users end their search after seeing an AI summary, vs without one

Then show that the clicks that do come through are worth more. Shopify’s Q1 2026 merchant data showed AI-referred visitors converting about 50% better than organic visitors and spending 14% more per order, with AI-referred orders up nearly 13x year on year. That’s ecommerce-specific, I know, but it’s a pretty good answer to ‘is there any real traffic?’

• ~50% higher conversion rate for AI-referred visitors than organic
• 14% higher average order value from AI-referred visitors
• ~13x more AI-referred orders year on year

Then show what you can actually measure:

• GA4’s AI Assistant channel, switched on by default since May, which reports traffic and conversions from ChatGPT, Claude, Gemini and others as their own line. (Here’s how to track AI traffic in GA4.)
• Google Search Console, which now reports appearances in AI Overviews and AI Mode.
• Prompt monitoring and server logs, which show citations, share of voice against named competitors and AI bot activity, month on month.

The honest caveat is that none of this is perfect yet. GA4 tends to undercount, because some apps strip the referrer, so the real AI traffic figure is often higher than it looks. Prompt tracking is a representative sample, not a complete record. I’d always rather be upfront about that with a client, because a reliable indicator presented honestly will earn more trust than a perfect-looking number that falls apart three months later.

Frame the whole thing as an investment in brand visibility and authority, with traffic as one of the payoffs. And remember: Muck Rack found 39% of PR professionals aren’t measuring GEO at all. The moment you start, you’re already ahead of a lot of the market.

3. ‘We’ll Wait and See’

This objection is really about risk, and the most useful thing I’ve learned in my years in marketing is that waiting is never actually free. Shopify has seen sessions referred from AI tools grow more than 8x year on year, against 5% growth in organic search. Competitors who start now are building up citations and authority, and a client who turns up in 18 months’ time will be starting from scratch against someone with an 18-month head start.

• 8x+ growth in AI-referred sessions, year on year
• ~5% growth in organic search sessions over the same period

I wouldn’t go as far as telling them the sky is about to fall in. Just be clear about what waiting actually costs them.

4. ‘We Don’t Have the Resource’

This one is about effort rather than risk, and I’d almost flip it round: that’s exactly why GEO is a managed service. The work sits with you (and your white label partner), so the client doesn’t need to hire anyone or build a team. That’s what agencies have been for since the beginning of time.

If neither answer lands in the moment, don’t force it. Run another visibility check a few weeks later. In our experience the gap rarely closes by itself, and seeing it hold steady, or get wider, with named competitors in it does more persuading than any pitch could.

How to Deliver GEO Without Building a Team

Which brings us neatly to the question I imagine some of you have been sitting on since the four pillars: if we sell this, who actually does the work, and do the numbers stack up for us? It’s the part everyone tends to dance around in these conversations, so I’ll keep it simple. There are three ways to work with SUSO:

Model	How it works	What’s in it for you
White label	You own the client relationship and the billing, and we deliver under your brand. We use alias emails by default, so the client never sees SUSO, only AI search performance that keeps improving.	Your own GEO service, without building a team, plus your margin on top of our cost
Referral	You pass the client to us.	A commission, with no delivery responsibility, and you still get to be the hero of the piece
Collaboration	We build the GEO service with you.	The gaps in your current offering, filled

On pricing, our service tiers run from around $1,000 to $3,600 USD a month for a full program, with no long-term commitment. On white label GEO, partner margins typically sit between 20% and 50% on top of our cost. We’ve built the numbers so they work for you, for us and for the client. Everybody eats.

The Bigger Picture: The Plates Have Already Moved

I did promise to spare you any more metaphors, and I’ve mostly kept my word. But I’m going to allow myself one last trip back to the tectonic plates, because they make the point better than I can.

The shift has already happened. Your clients are on the other side of it, asking ‘what should we do and why?’, and the gap between them and the answers is exactly where Agency Land gets to prove its worth. The agencies that win in 2027 won’t be the ones with the cleverest theory of GEO. They’ll be the ones who started having these conversations this year.

You don’t need to become a GEO specialist overnight. You need the framing, the data and a partner who can fill the gaps, which is exactly what the SUSO Partner Club is for:

• The Partner Club Hub: one place for a white label pitch deck, proposal template, objections guide, intake form, report templates and case studies. Members get first access.
• The free AI Search Visibility Checker: a five-minute snapshot of any brand’s AI visibility.
• Free workshops and training, including our Writing Content for AI Search workshop.
• Our specialist helpdesk, with detailed answers to tricky client questions within 24 hours.

The Partner Club is free to join. And if you want any of these assets before the Hub launches, just get in touch with Cara – there’s no need to wait.

Want the full session? Watch the webinar recording and read the transcript here.

FAQs

• How do agencies sell GEO to clients?

  Lead with the problem, not the solution. Run a visibility check, show the client where competitors appear in AI answers and they don’t, explain what’s causing the gap, then propose a plan built on the four pillars: prompt monitoring, earned/PR, on-site content and technical. Frame the price against what a new customer is worth.

• What’s the difference between SEO and GEO?

  SEO aims to rank pages in search results. GEO aims to get a brand cited in AI-generated answers. They share foundations like quality content, E-E-A-T and crawlability, but GEO adds prompt research, AI bot access and mentions on the sources LLMs favor, such as Reddit, Wikipedia and review sites.

• How long does a GEO audit take?

  Typically two to three weeks. Prompts need to be tracked for at least two weeks to give a reliable reading, and they can’t be tracked retroactively. A full audit also covers content, technical factors and the wider web.

• How much does GEO cost?

  SUSO’s GEO service tiers range from around $1,000 to $3,600 USD a month for a full program, with no long-term commitment. Agencies reselling on a white label basis typically add a 20-50% margin.

• Can you measure ROI from GEO?

  Yes, and increasingly well. GA4 has a default AI Assistant channel, Google Search Console reports AI Overview and AI Mode appearances, and prompt monitoring tracks citations and share of voice against competitors. GA4 tends to undercount AI traffic, because some apps strip the referrer.

• Which industries see the most success with GEO?

  Direct-to-consumer, lifestyle and consumer goods brands tend to see strong results, especially anything that appears in ‘best of’ listicles. But every industry has the LLMs that matter most to it. In financial services, for example, Perplexity carries particular weight.

• Is there still a pitch if a client already performs well in AI search?

  Yes. Models keep training on new data, competitors keep investing, and platform updates can shift citation patterns quickly. Ongoing monitoring shows whether a strong position is holding or slipping, and a three-month comparison is often what reopens the conversation.

• Does owned content or earned media matter more for AI visibility?

  It depends on the industry and the prompt set. Earned media is highly influential, because consistent third-party coverage expands a brand’s footprint across the web. Owned content still matters, especially product and collection pages pulled into AI shopping results. Above all, owned and earned messaging must be consistent, because LLMs cross-check claims across sources.

• How quickly do LLMs pick up new content?

  Sometimes within 24 hours. We’ve seen direct correlations with LinkedIn Pulse articles, for example. Freshness matters as much as it ever did. The difference is that models can now read content from yesterday, not from six months ago.