By Steve Merrill, Founder of WRKNG Digital | June 15, 2026
Most agencies claiming AEO expertise didn't develop it. They renamed a service deck.
That matters a lot if you're a Shopify store owner trying to figure out who ChatGPT, Perplexity, or Google's AI Overview actually trusts when someone asks what product to buy. The wrong agency burns your budget running the same technical SEO playbook with "AI" bolted onto the proposal. The right one understands that AI systems index content differently than search engines do, and builds your visibility accordingly.
I've spent the last eight months building AI visibility tools specifically for Shopify stores. We've audited hundreds of product catalogs and analyzed how AI assistants decide what to recommend. What I keep seeing from agencies is troubling. Rebranded deliverables. No citation tracking. Schema audits that check the same boxes they checked in 2019.
This post is about telling them apart.
What Does AEO Actually Involve in 2026?
Answer Engine Optimization is the work of making your content, product data, and site structure legible to AI systems so they pull from you instead of a competitor. It's not a subset of SEO. The mechanics are different.
Google ranks pages. AI systems synthesize answers from sources they've already determined to be trustworthy and well-structured. A page can sit at position one on Google and still never appear in a ChatGPT or Perplexity response. According to Search Engine Land's analysis of Google AI Overview sourcing patterns, the top-ranking organic result appears in an AI Overview only about 12% of the time.
Real AEO work in 2026 covers four things: structured data that AI can parse cleanly, FAQ content written to answer discrete questions (not to rank for broad keywords), citation tracking that measures whether your brand actually appears in AI-generated responses, and content structured for extraction rather than for scroll-through reading. If an agency can't explain all four, they're not doing AEO.
What Are the Three Questions Every Agency Should Be Able to Answer?
Ask these before you sign.
First: "How does AI index content differently than Google, and how does your work account for that?"
A real answer gets specific. AI systems favor content that directly answers a discrete question in the first two sentences. They favor schema markup that labels entities, products, and organizations clearly. They favor sites with consistent, verifiable entity signals across multiple platforms. A rebranded SEO agency will give you something vague about "quality content" and "technical health." Those aren't wrong answers. They're just not answers to the question you asked.
Second: "How do you track whether a client appears in AI-generated responses?"
Citation tracking is its own discipline. You're asking whether the agency actively tests AI responses across tools like ChatGPT, Perplexity, and Copilot for your target queries, logs those results over time, and ties that data to the content work they're doing. Most agencies don't do this at all. They report organic traffic and call it a win.
Third: "Can you show me a case study where a client's AI citation rate improved?"
Real AEO case studies show before-and-after citation data. Not "rankings improved" or "traffic went up." Those metrics can move for dozens of unrelated reasons. AI citation improvement is specific. It shows the agency actually knows what they changed and why it worked.
What Red Flags Should You Watch For?
Walk away fast. These are the signals that tell you an agency is faking it.
They lead every conversation with keyword rankings. Google rankings are a real metric. They're just not what you're hiring an AEO agency to move. If rankings are the first KPI they mention, they're running an SEO engagement with AEO branding.
They can't describe how AI platforms pull content differently from web crawlers. This is foundational. Bing, ChatGPT, and Perplexity each have different content ingestion behaviors. Semrush's research on AI search ranking factors shows that entities, FAQ structure, and E-E-A-T signals matter in AI responses in ways that differ from classic Google ranking factors. An agency that can't explain this at a basic level hasn't done the reading.
They propose schema work without auditing what you already have. Schema is one of the most concrete AEO levers available. But just adding JSON-LD without auditing for conflicts, missing properties, and incorrect entity relationships doesn't help. It can actively confuse AI indexers if it contradicts what's on the page.
They have no process for tracking AI mentions. This is the clearest tell. If they can't show you how they measure AI citation performance, they're not running an AEO program. They're running something that sounds like one.
What Does Real AEO Work Actually Look Like?
Concrete and unglamorous. Here's what the actual deliverables should include.
Schema work goes beyond Product and Organization markup. It covers FAQPage, HowTo, ItemList, and BreadcrumbList where applicable. A proper audit maps what you have, what's missing, and what AI systems are actually reading versus ignoring.
FAQ content gets built around actual user questions. The goal is a direct answer in the first sentence. AI systems extract the first clean answer they find. Every FAQ entry should pass this test: if someone asked this question to an AI assistant, would this answer appear verbatim? If not, the content isn't written for extraction.
Citation monitoring tracks real AI responses. Run target queries against ChatGPT, Perplexity, Copilot, and Google's AI Overview on a regular cadence, log whether your brand appears, and analyze whether citations come from your product pages, your blog, or a third-party source. Search Engine Journal's AEO guide outlines the citation tracking framework most serious practitioners use as a starting point.
Content restructuring moves answers above the fold. Long-form content doesn't disappear, but the structure changes. Answers lead. Detail follows. Section headers become questions. Paragraphs get shorter and more discrete. That's a different writing discipline than traditional content marketing.
How Do You Read an AEO Agency's Case Studies?
Look for specificity. Vague case studies are the clearest sign of vague work.
A real AEO case study names the type of structured data added. It shows AI citation rate before and after. Traffic data tells you nothing specific about AI visibility. The case study identifies which AI platforms began citing the client, on what query types, and describes exactly what content changes addressed which extraction gaps.
I've run this evaluation on my own clients' audits. What I see again and again is that the stores showing up in AI responses didn't have better products. They had better data. More complete schema. FAQ pages written as direct answers. Product descriptions with entity signals that AI systems recognize and trust.
What's the Difference Between Ranking on Google and Being Cited by ChatGPT?
Ranking is about relevance signals for a crawl-and-index system. Citations are about trust and extractability for a synthesis system.
Google rewards pages it can rank relative to other pages for a given query. ChatGPT and Perplexity reward sources they can extract a clean, trustworthy answer from. You can win one without winning the other.
The brands that are getting cited right now did something specific: they wrote content that answers a question in one paragraph, labeled their entities clearly with schema, and built enough off-site authority that AI systems treat them as reliable sources. That's the formula. An agency that hasn't built that formula for a client in production hasn't done AEO.
Which Agencies Are Actually Doing This Work?
Short list. Honest framing.
A handful of agencies have been doing genuine AI search optimization work rather than renaming SEO. NP Digital has published substantive research on AI Overview sourcing. iPullRank has run technical experiments on entity optimization and AI content extraction. Terakeet has built entity-authority frameworks that predate the AI search wave and translate well to AEO. Conductor has added AI citation features to its platform and its agency work reflects that. Siege Media produces structured content built specifically for AI extraction.
Beyond those five, you're evaluating on a case-by-case basis. Ask the three questions above. Review case studies for specificity. Ask to see how they'd approach one of your product categories before signing a contract. The agencies doing real work welcome that conversation. The ones rebranding from SEO will stall or pivot to rankings.
Five more worth researching: Wpromote, Ignite Visibility, Online Optimism, Clearlink, and Directive Consulting. None of these are endorsements. Do your own evaluation using the three questions above.
Real AEO isn't a mystery. Disciplined work. Measurable outputs. The agencies that know this welcome scrutiny.
Frequently Asked Questions
What is the difference between AEO and SEO?
SEO targets search engine ranking algorithms that crawl, index, and rank pages by relevance signals. AEO targets AI synthesis systems that extract answers from trusted, well-structured sources and present them in generated responses. The two disciplines share some technical foundations (schema, E-E-A-T, content quality) but differ in how they measure success and what they optimize toward. An SEO strategy won't automatically produce AI citations. That requires additional, specific work.
How do I know if an AEO agency is legitimate?
Ask three things: how they explain the difference between AI indexing and traditional web crawling, how they track AI citation performance across tools like ChatGPT and Perplexity, and whether they can show a case study with citation-rate data. Agencies doing real AEO work can answer all three specifically. Those rebranding from SEO will give vague or ranking-focused answers.
What does AI citation tracking actually involve?
Running target queries against ChatGPT, Perplexity, Google's AI Overview, and Microsoft Copilot on a regular schedule and logging whether your brand, products, or content appear in the generated responses. You track which AI platform cites you, on what types of queries, and whether citations link to your own content or third-party mentions. This is the core measurement discipline of AEO work.
Does structured data actually help with AI citations?
Yes, but it has to be correct structured data. Schema markup that labels your products, FAQs, organization, and content entities clearly helps AI systems parse and trust your information. Incorrect or conflicting schema can do the opposite. A schema audit should always come first.
Can a Shopify store do AEO without hiring an agency?
Some of it. You can add FAQ schema to product pages, rewrite product descriptions to answer specific questions in the first sentence, and run manual citation checks against AI assistants. The parts that get harder without dedicated tools are citation tracking at scale, entity optimization across a large catalog, and systematic content restructuring. The Shopify stores making real gains in AI visibility are doing the systematic version.
Want to see how your Shopify store shows up in AI search right now? We audit product feeds, structured data, and content structure to show you exactly what AI assistants see when someone asks about your category. Get your AI visibility audit from WRKNG Digital.
