This is a piece written by an agency about how to choose an agency. Be skeptical of it.
The genre, “5 questions to ask your X agency,” is one of the most dishonest forms of marketing content. It’s almost always five questions the publishing agency happens to be excellent at answering. By the time the reader finishes, the only agency that looks like a fit is the one that wrote the piece.
Ima try to do better.
There are real ways to evaluate a Generative Engine Optimization agency. Real questions worth asking, and real cases where the right answer is “don’t hire one yet.” I’ll cover all of it, including the situations where GNW Consulting (the firm I work for) is the wrong choice.
If you finish this piece convinced you should hire us, great. If you finish it convinced you should hire someone else or do this in-house, less great, but I understand. The goal is that you make the right call for your situation. Before you dive in, the 2026 State of GEO report gives you the market context that makes everything below more useful.
Here’s the order:
- Why this is hard to evaluate in 2026.
- Do you actually need an agency?
- The diagnostic for telling real GEO agencies from rebranded SEO agencies.
- The questions worth asking.
- When to walk away from agency offers entirely.
Why this is hard to evaluate in 2026
The single biggest reason it’s hard to choose a GEO agency right now: almost every agency claims to be one.
In a survey of 225 B2B marketing and revenue leaders we ran in Q1 of this year, 88% reported that their existing SEO agency now claims to offer GEO or AI search optimization services. 37% admitted those services are “loosely defined.”
A small slice of those agencies are genuinely investing in the new capability set. The rest have added “GEO” and “AI-optimized” to their service descriptions without changing what they actually do.
This isn’t malicious. Most SEO agencies are responding rationally to client demand. Clients ask about GEO, so the service description gets updated. But “we offer GEO services” can mean anything from “we built a dedicated practice with new tools and people” to “we tweaked our content briefs to include a few additional schema fields.” Both versions show up on the website. The reader can’t tell which is which without doing real diligence.
So you can’t take service-page claims at face value. You have to verify everything, and the verification is harder than it sounds because most agencies have learned the right vocabulary even when they haven’t done the work.
The good news: the verification is doable. The diagnostic below will give you most of what you need.
Step 1: Decide whether you actually need an agency
Before you evaluate a single agency, decide whether hiring one is the right move at all. Three honest cases where it isn’t.
Case 1: You have an in-house team with the bandwidth and the curiosity, and you’re early in your GEO journey. If you have a marketing operations or content lead with genuine interest in this discipline, and they have the time to dedicate 10-15 hours per week to learning and executing, you can do good GEO work in-house in the first 12 months. The tools are accessible. The methodology is available (including in pieces like this one, and in the foundational thinking we’ve published on what GEO actually is and why buyer prompts have to come first). What you’ll lose by going in-house is the benefit of an agency seeing across many clients — but you’ll learn faster by doing the work yourself than you would by managing an agency that’s doing it for you.
Case 2: Your competitive landscape is small and stable. If your category has 3-5 competitors and none of them are investing in GEO yet, the cost of being first is low. You can probably start with a basic in-house effort — clean up your content structure, build a small Reddit and LinkedIn presence, get a few G2 reviews refreshed — and capture meaningful share without paying agency rates. Agencies become more valuable when the competitive landscape is dense and the optimization gets nonlinear. If you’re in a quiet category, do it yourself.
Case 3: You’re not ready to act on what an agency would tell you. This is the most important one and the least discussed. If your team can’t get budget approved, can’t turn content around in reasonable time, can’t ship changes to the site quickly, can’t get sales to show up for customer case studies, then an agency engagement produces a beautiful strategy document and zero results.
Agencies amplify functional teams. They cannot fix broken ones. If your team can’t execute against your current strategy, an agency won’t fix that. Invest the budget on team-building instead.
If none of those three cases apply to you, you might benefit from an agency. Read on.
Step 2: The diagnostic for telling real GEO from rebranded SEO
This is the most important section. The standard advice for evaluating any agency is “ask to see their work.” That advice is right for SEO agencies, where the work is high-volume and visible — backlinks, content briefs, technical audits, ranking reports. It’s the wrong test for GEO agencies right now, because GEO in 2026 isn’t yet a high-volume execution discipline. It’s an insight-and-adjustment discipline.
The best GEO agencies right now aren’t producing five-times-as-much-content. They’re monitoring citation patterns across clients and platforms, spotting which adjustments correlate with movement, and bringing those insights back to client teams that do the execution. The agency’s value is the cross-client pattern recognition you can’t build in-house, not the deliverable volume.
So the test isn’t “show me 90 days of Reddit posts you wrote.” The test is some version of: what have you been seeing, and how does that translate into what we should adjust?
Here’s how to run that test.
Ask them what they’re currently seeing in your category — or in categories adjacent to yours. A real GEO agency monitors citation patterns continuously across clients and across platforms. They should be able to tell you, with some specificity: which sources are getting cited most often for prompts in your category right now, how that’s changed in the last few months, and what’s different about your category compared to others. If they can’t, they’re not running the monitoring discipline that makes the agency useful in the first place.
Ask them to walk you through what they monitor and how. Specifically: what citation-tracking tools they use (Profound, LLMO by Adobe, and a few others are the current options), how often they pull data, what they look for, what counts as a meaningful shift versus normal noise. If the answer is “we look at AI search visibility scores in our reporting tool,” you’re talking to an agency that’s consuming data, not producing insight. The agencies actually doing GEO well have a point of view on the limits of these tools and a methodology for interpreting them. (We wrote about what we actually know and don’t know about LLM citation signals — that piece is roughly the level of epistemic honesty the agency should bring.)
Ask them to describe a recent insight that changed what a client did. Not a case study with bragging numbers — a story about something they noticed that led to an adjustment. “We saw that a competitor was getting cited disproportionately for a specific prompt type, traced it back to a piece of Reddit content, and recommended the client build their own community presence around that thread pattern.” That kind of story tells you the agency translates monitoring into adjustments. Vague answers (“we help clients optimize their AI search visibility”) tell you they’re describing capability instead of demonstrating it.
Ask them what they don’t know. This is the most diagnostic question on the list. A real GEO agency will tell you specific things: that signal weighting differs by platform and shifts over time, that universal ranking factors don’t exist, that visibility scores are directional rather than precise, that Claude attribution is structurally broken because it doesn’t surface citations in its base UI. They’ll be comfortable saying “we don’t know exactly how X works yet — here’s how we handle that gap.” An agency that confidently answers everything is overclaiming, and an agency overclaiming in the evaluation conversation is going to overclaim in the engagement.
The five surfaces test, slightly different from how it’s usually framed. GEO scope includes work on five surfaces beyond your website: community platforms (Reddit, forums), video (YouTube primarily), peer-review sites (G2, TrustRadius), LinkedIn company-page content, and citation tracking across the major LLMs. A real GEO agency has a view on all five — they monitor them, they form opinions about which matter most for your category, and they bring those opinions to your team. They might or might not be producing deliverables on all five. (Honestly, in 2026, nobody is producing high-volume execution on all five.) But they should be able to tell you what they’re seeing on each one and where the highest-leverage adjustments are.
The shift in framing is the point: you’re not hiring an agency to do twice as much content. You’re hiring them to see across more data than your team can see, and to translate that into a small number of high-leverage adjustments your team makes. The execution probably stays with your team in most engagements right now. The judgment is what you’re buying.
If the agency you’re evaluating can describe what they monitor, give you a recent insight that led to a real client adjustment, and name the things they don’t know — they’re doing GEO. If they can only describe deliverables and capability, they’re describing themselves as if GEO were a high-volume SEO discipline. It isn’t yet. Be cautious.
Step 3: The questions worth asking beyond the diagnostic
If an agency clears the insight-based diagnostic above, they’ve passed the bar of actually doing GEO. The remaining questions are about whether they’re a fit for your specific situation.
“How do you start an engagement?” Listen for whether they start with the buyer or with the website. A buyer-first agency will tell you about prompt mapping, customer interviews, and figuring out which AI tools your buyers actually use. A website-first agency will tell you about a technical audit and a content gap analysis. Both have value. But buyer-first work is what differentiates GEO from SEO, and an agency that doesn’t lead with the buyer is probably executing GEO with an SEO operating model. (We wrote more about why this matters in the buyer-prompt methodology piece.)
“What signals do you optimize for?” Listen for hedged language. An agency that confidently lists “the top 5 signals that drive LLM citations” is overclaiming — the honest answer is that the signals depend on your buyer, your category, your stage of funnel, and the specific platform, and they shift over time. (We wrote about what we actually know and don’t know about LLM signals.) Look for an agency that names uncertainty and describes a measurement process, not one that promises a deterministic ranking factor.
“What does your measurement look like?” The honest answer is some version of: AI referral traffic is measurable and reliable; GEO visibility scores are directional and useful for spotting patterns; LLM citation tracking gives you competitive intelligence but isn’t precise performance reporting. If an agency promises precise visibility metrics they can tie directly to pipeline, ask harder questions. The current state of the measurement infrastructure doesn’t support that level of precision yet.
“Show me one client you’ve fired, or one that fired you, and tell me what went wrong.” This is the most useful question on the list. Every agency has had relationships that didn’t work. An agency that gives you a real answer — “we had a client who needed daily content velocity we couldn’t sustainably provide, and we recommended they hire a content shop instead” — is showing you their judgment about fit. An agency that says “we haven’t really had that situation” is either lying or new.
“How do you handle the surfaces where my buyer isn’t yet?” GEO is partly about showing up where your buyer is, and partly about building presence in advance on surfaces your buyer might use later. A good agency has a view on the second part. A weak agency only talks about the first.
“What’s your team makeup?” A GEO team needs at minimum: someone who understands content and citation structure, someone who understands community platforms (and is comfortable on Reddit specifically), someone who can produce or coordinate video, someone who reads analytics critically, and someone with strategy chops to tie it together. If the agency you’re evaluating has three SEO specialists and an account manager, they’re going to struggle on the community and video surfaces no matter how good their content work is.
Step 4: When to walk away from agency offers entirely
A few clear “walk away” signals:
They promise specific rankings or visibility numbers. Nobody can guarantee LLM citation positions in the way some SEO agencies could (mostly dishonestly) guarantee Google rankings. If the agency commits to “we’ll get you to top-3 in ChatGPT for your category in 90 days,” they’re either inexperienced or selling something that won’t hold up.
They have a “GEO playbook” they apply to every client. Buyer-first methodology means the work is necessarily different for every category. An agency selling you the same playbook they sold the last 20 clients is doing one of two things: either they’ve found a generalizable framework (rare but possible) or they’re just running the same SEO process on everyone with new vocabulary (much more common).
Their pricing doesn’t account for the off-site work. A retainer that covers website optimization but charges separately for “community management” or “video work” or “review-site campaigns” is structurally pricing the agency to focus on the easy work. The off-site surfaces are where GEO differentiates from SEO. If those are extras, the engagement will skew toward the on-site work even if you want it to be balanced.
They can’t describe a GEO-specific insight separately from SEO work. Ask: “Walk me through something you’ve seen in citation patterns or LLM behavior that wouldn’t have shown up in traditional SEO monitoring.” If the answer is “well, our SEO work also helps with GEO because of structured data,” you’re not talking to a GEO agency. You want to hear something the agency has noticed that only shows up when you’re watching LLMs as the primary signal — a citation shift, a competitor pattern, a platform-specific behavior. If they can’t separate the two disciplines in their answer, they probably can’t separate them in their work either.
The honest pitch
GNW Consulting does GEO work. We’re one of the agencies you might consider hiring. If everything I’ve written above sounds like an accurate description of the discipline and a fair framing of the buyer’s evaluation problem — that’s because I tried to write it as if you weren’t going to hire us, and let the diagnostic stand or fall on its own.
If you’ve gone through the diagnostic and you’re considering us along with two or three other firms, we’d encourage you to ask us the same questions you ask them. What we’re seeing in your category right now. What we monitor and how. A recent insight that led to a client adjustment. What we don’t know. The “show me a client you fired” question. We’ll answer them honestly, including the cases where we wouldn’t be a fit.
If you want a low-commitment first step, we run free competitive GEO audits that show you how your brand currently shows up in ChatGPT, Perplexity, and Google’s AI Overviews compared to your competitors — including specific recommendations for what to fix first. It’s useful regardless of whether you eventually hire us, another agency, or do the work in-house. Audit delivered in 5-7 days. No commitment, no sales call unless you want one.
Frequently Asked Questions
What is the difference between SEO and GEO?
Search Engine Optimization (SEO) focuses on ranking websites in traditional search results like Google. Generative Engine Optimization (GEO) focuses on increasing a brand’s visibility and citation frequency within AI-powered answer engines like ChatGPT, Perplexity, and Google Gemini. While SEO prioritizes keywords and backlinks, GEO prioritizes information density and cross-platform authority.
How do I track GEO rankings in 2026?
Standard keyword trackers do not work for GEO because AI responses are personalized and dynamic. To track performance, brands use specialized tools like Profound or LLMO by Adobe to monitor citation rates and Brand Share of Voice in AI overviews. The most reliable metrics include direct referral traffic from AI engines and consistent presence across non-website surfaces like Reddit and G2.
Do I need a specialized GEO agency?
You may not need a specialized agency if you have a high-functioning in-house content team with the bandwidth to experiment. However, a specialized GEO agency is valuable if you operate in a highly competitive market where pattern recognition across multiple clients is required to maintain a digital share of voice.
Does traditional SEO help with GEO?
Yes, foundational SEO elements like technical site speed and structured data (schema markup) help AI agents crawl and understand your content. However, traditional SEO alone is insufficient for GEO because it ignores the off-site signals—such as community discussions and peer reviews—that LLMs weigh heavily when generating answers.
What should a GEO audit include?
A thorough GEO audit should analyze how your brand appears in ChatGPT, Perplexity, and Gemini compared to your top competitors. It should identify “citation gaps” where competitors are being sourced instead of you and map out specific buyer prompts that are currently failing to trigger your brand as a recommended solution.
Related reading from GNW:
What Generative Engine Optimization Actually Is — And Why the Term Matters — the canonical definition piece.
Dentists Don’t Prompt ChatGPT Like Marketers Do — on the buyer-first methodology that should anchor any GEO engagement.
What We Actually Know About LLM Citation Signals — And What We Don’t — on calibrated uncertainty in a discipline where confident rankings are mostly wrong.
Are SEO and GEO Really the Same? — short video walkthrough of the six core differences.
Questions to Ask Your SEO Agency — short companion video to this piece.
