Answer engine
optimization for
commerce brands.

For twenty years, the goal of getting found online was simple to state: rank on the first page of Google. The whole discipline of SEO was built around that single objective. A customer typed a query, Google returned ten blue links, and you fought to be one of them. That world is ending, not slowly, and most commerce brands have not adjusted their playbook to the thing replacing it.

The thing replacing it is the answer engine. A growing share of buyers no longer scan a page of links. They ask ChatGPT, Perplexity, Google's AI Overviews, Gemini, or Claude a question and read a synthesized answer that names a few products or brands and moves on. AI-driven search jumped from under 10% of interactions in 2023 to roughly 30% by 2026, and ChatGPT alone now reports over 400 million weekly active users (Position Digital). The query did not disappear. The ten blue links did. Your brand is either inside the answer or it is invisible.

Answer Engine Optimization, AEO, is the discipline of becoming the answer. It overlaps with SEO and shares DNA with what I covered in the GEO versus SEO piece, but it is its own practice with its own mechanics, and the brands that learn it early will own the new shelf the way early SEO winners owned the old one. This is the operator's guide: how answer engines actually decide what to cite, the levers that get you in, the structured data and off-site work that commerce specifically needs, and a 30/60/90 plan to start.

The new acronyms,
in language a
human uses.

This field invented a pile of new terms in about eighteen months, and a lot of them describe the same thing from slightly different angles. Here is the honest map before we go deep.

One clarification worth making up front: AEO and GEO are, for practical purposes, the same job. Different agencies brand it differently. I will use AEO throughout and treat GEO as its twin. What matters is the work, not the acronym.

This is not a
trend. It is a
channel forming.

It is easy to wave this off as hype, so start with the size of it. AI chatbot sessions have been roughly doubling every year since 2024, reaching well over a billion monthly visits by 2026 (Position Digital). The share of search-like interactions happening inside an answer engine rather than a traditional results page crossed the point where it can no longer be ignored. And critically, the engines are fragmenting. The reflex is to treat this as a ChatGPT story, and for a while it was: one report had ChatGPT driving 87% of measurable AI referral traffic. By early 2026 that concentration had loosened, with ChatGPT's share of measurable AI referrals falling to around 63%, Claude rising near 18%, Gemini around 11%, and Perplexity around 7% (Goodie).

That fragmentation matters for how you work. You are not optimizing for one engine, you are optimizing for a category of them, and they do not all cite the same kinds of sources. Optimizing only for ChatGPT is the AEO equivalent of optimizing only for one search engine in 2010. The winners build for the mechanics that are common across engines, then tune for the differences.

What actually
earns a citation.

Answer engines are not magic, and they are not random. Researchers have started to reverse-engineer what gets content pulled into an answer, and the findings are consistent enough to act on. The foundational work here is the Princeton and Georgia Tech GEO study, which tested specific content changes against real generative engines and measured the lift (Princeton/Georgia Tech, GEO). A few patterns hold up across studies.

Factual density wins. Engines disproportionately cite content packed with specific numbers, percentages, and quantified claims over content that is all adjectives and opinion. The Princeton work found that adding relevant statistics to a page lifted its visibility in AI answers by 30% to 40%. A separate analysis of 10,000 queries found that pages with structured lists, direct quotes, and statistics had 30% to 40% higher visibility (Search Engine Land). The lesson is blunt: vague brand copy does not get cited. Specific, sourced, numerical content does.

Freshness wins. Recency is weighted heavily. One analysis found 85% of AI Overview citations came from content published within the last two years, with 44% from the most recent year alone, and that recently updated content appears several times more often than stale material (ALM Corp). For commerce, this is a gift, because your category changes constantly and you have a legitimate reason to refresh.

Different engines trust different sources. This is the part most brands miss. ChatGPT leans heavily on Wikipedia-style authority, which accounts for a large share of its top citations. Perplexity leans toward Reddit and community discussion. Google's AI Overviews balance professional and social sources. So the off-site work that earns you a citation in one engine is not the same as another, which we will get to.

Five levers, in
order of impact.

The mechanics resolve into a short list of things you actually do. Here they are, roughly ranked by how much they move citation rates for a commerce brand.

Notice the shape of this list. The first three are on-page and within your control. The fourth is off-page and slower. The fifth is technical and binary, either the door is open or it is not. Most brands obsess over content and never check whether the crawlers can even reach it. Start at the bottom of the list and work up: confirm access first, then build the content that deserves to be found.

Answer-first, in practice

The single highest-return change for most commerce content is restructuring it answer-first. Lead every important page and section with a direct, self-contained answer of forty to sixty words, then expand below it. An engine scanning your page for a usable quote wants a clean, liftable statement, not a paragraph that buries the point under three sentences of preamble. This is the same instinct that makes a good product page convert, which I broke down in the product page audit: say the thing, then support it.

Make your catalog
machine-legible.

Structured data is no longer optional for a commerce brand that wants to be cited, because it is how an answer engine reads your catalog without guessing. The schema types that matter most for commerce, in rough priority:

Product schema is the foundation. It tells the engine the product name, price, availability, brand, GTIN, and ratings in a format it does not have to infer from your layout. A product page without it is asking the machine to reverse-engineer what the page is from messy HTML, and the machine will often get it wrong or skip it. Review and AggregateRating schema exposes your social proof as data, which matters because reviews feed the engine's sense of whether you are a trustworthy answer, a point I made about visibility in the ChatGPT visibility piece.

FAQPage schema is the quiet overperformer. Pages marked up with FAQ schema are reported to be several times more likely to appear in AI Overviews than pages without it, and they show meaningfully higher citation rates across engines (ALM Corp). The reason is structural: an FAQ is already a question paired with a clean answer, which is exactly the shape an answer engine wants. Build genuine FAQ sections on your category and product pages and mark them up. Organization schema ties it all to a clear brand entity, with your logo, social profiles, and sameAs links, so the engine knows who you are and connects the mentions of you across the web to a single recognized brand.

What others say
about you matters
as much as your site.

Here is the part that frustrates brands used to controlling their own pages: a large share of what gets you cited happens off your website entirely. Answer engines build their picture of you from the whole web, and they weight independent sources heavily because independent sources are harder to game. This is where the engine-by-engine differences become actionable.

ChatGPT leans on encyclopedic, authoritative sources, with Wikipedia-style references making up a large share of its top citations. The implication for a brand is to earn the kind of structured, factual presence those sources contain: a clean Wikipedia entry if you genuinely qualify, accurate listings in industry databases, mentions in established publications. Perplexity, by contrast, leans hard on community discussion, with Reddit a dominant source. That means real participation and genuine mentions in the communities where your category is discussed are worth more for Perplexity visibility than another blog post on your own domain. I made the broader case for community channels in the Reddit as a commerce channel piece, and AEO is another reason that channel is underrated.

The honest version of this lever is that it is slow and cannot be faked at scale. You earn third-party authority by being genuinely talked about, reviewed, and referenced, which is a function of having a product and a point of view worth talking about. There is no schema tag for that. But you can accelerate it: seed honest reviews, participate in the right communities as yourself, pitch real data to publications, and make sure that when someone does mention you, your own entity is clear enough that the engine connects the dots back to you.

None of it counts
if the door
is locked.

The least glamorous lever is the one that quietly disqualifies brands before they start. Answer engines read the web through named crawler bots, and if your site blocks them, you cannot be cited no matter how good your content is. The major ones to allow are GPTBot for ChatGPT, ClaudeBot for Claude, PerplexityBot for Perplexity, and Google-Extended for Google's AI products. Check your robots.txt and your CDN or bot-management settings, because plenty of brands block these by default or via an aggressive security rule and never realize it.

Beyond access, the same technical hygiene that helps human SEO helps machines: fast load times, clean HTML, a logical heading structure, and a current sitemap. A slow or messy page is harder for an engine to parse and less likely to be used, which is one more reason store speed is a revenue issue, as I argued in the store speed piece.

On llms.txt, the proposed file that points AI at your important pages: be calm about it. As of early 2026, no major AI provider has confirmed using it as a ranking or citation signal (ALM Corp). It takes about thirty minutes to add and is reasonable insurance against the chance it matters later, but it should sit at the bottom of your list, not the top. Anyone selling llms.txt as the key to AI visibility is selling the easy part because the hard parts are hard.

Your product is
the answer to a
buying question.

Most AEO advice is written for B2B and content sites, where the answer is an article. For commerce, the answer is often a product, and that changes the work. When a buyer asks an engine "what is the best insulated water bottle for hiking," the engine is going to name a few products. The question is whether yours is one of them, and that comes down to how legible your product is as an entity: clear product schema, genuine reviews exposed as data, a product page that states plainly what the product is and who it is for, and third-party corroboration that it is good.

There is a measurement problem you should understand going in. Much of this channel is a dark funnel. A buyer asks ChatGPT, gets your product named, and either buys inside the chat or searches your brand directly later, which shows up in your analytics as direct or branded traffic rather than as an AI referral. So the channel will look smaller in your dashboard than it actually is. I went deep on tracking this and the attribution gap in the ChatGPT visibility piece, and the practical takeaway is to track AI referrals where you can, watch branded search and direct traffic for lifts, and resist the urge to judge the channel only by the referrals you can cleanly attribute.

Where to start,
in order.

AEO is large enough to feel paralyzing, so here is the sequence I would run for a commerce brand starting today. It front-loads the cheap, high-impact technical and structural work and pushes the slow authority-building into the background where it belongs.

Days 1 to 30: Access and structure

Confirm the AI crawlers can reach you, GPTBot, ClaudeBot, PerplexityBot, and Google-Extended, by auditing robots.txt and your bot settings. Implement Organization schema sitewide and Product plus AggregateRating schema on every product page, then verify the output rather than trusting the app. Restructure your highest-intent pages answer-first, with a clean forty to sixty word answer up top. This is the month where you fix the things that silently disqualify you.

Days 31 to 60: Content and FAQ

Build genuine FAQ sections on your category and top product pages and mark them up with FAQPage schema, targeting the actual questions buyers ask about your category. Audit your best content for factual density and add real, sourced statistics, ideally your own data, since original numbers are the most irreplaceable thing you can publish. Refresh your priority pages so they read as current, because freshness is weighted heavily.

Days 61 to 90: Authority and measurement

Start the slow off-site work: identify the communities and publications where your category is discussed, participate honestly, and pitch your original data to anyone who covers your space. Make sure your brand entity is clean everywhere it appears. And stand up measurement, tracking AI referrals where you can and watching branded and direct traffic for the lifts that signal dark-funnel impact. By day ninety you have closed the technical gaps, built extractable content, and started the authority flywheel that compounds from there.

The shift from links to answers is the biggest change in commerce discovery since mobile, and it rewards the same thing it always has, just measured by a new judge. Be specific. Be current. Be verifiable. Make your catalog legible to a machine and corroborated by the world outside your website. The brands that treat AEO as a real discipline now will be the named answer when a buyer asks, and the ones that wait will be the brands the engine has never heard of. The shelf is being rebuilt. Get on it before it sets.

Getting found in the answer engines is becoming as important as ranking on Google ever was. Building that visibility is part of the DTC brand consulting practice, and the form takes two minutes: start the conversation.