Most businesses approach answer engine optimization the way people approach a new gym: with a large purchase and a vague plan. Subscriptions, dashboards, a consultant who says “entity” a lot. Then, three months later, someone asks an AI assistant what the company does and gets a confident paragraph about a different company with a similar name.
There is a cheaper first move. Before strategy, before content, before tooling, you run an entity test: a short, structured look at whether machines can tell who you are, what you do, and where the official version of you lives. It takes about ten minutes for a small business. It costs nothing. And it routinely surfaces the problems that no amount of clever content will fix later.
This article walks through the test, what each part is really measuring, and how to read the results like a practitioner instead of a tourist.
What is an entity test, and why does it come before AEO strategy?
Search engines and AI assistants don’t just match words anymore. They try to resolve things: a business, a person, a product, a place. An entity is a thing a system can identify consistently across sources. If your business is a well-formed entity, a model can connect your website, your profiles, your mentions and your structured data into one coherent picture. If it isn’t, the model guesses, and guesses are where hallucinated founders and invented service areas come from.
Strategy assumes the entity exists. The entity test checks whether it does. Running it first is like checking that a building has a foundation before debating the paint.
What does the test actually check?
The Free AI-Visibility Quick Check is a one-page, printable version of this test. Its ten checks group naturally into three questions:
- Can machines identify you? Consistent naming everywhere, a clear About page, and Organization schema with sameAs links to your official profiles.
- Can machines reach and read you? Crawler access for the bots you want to allow, and pages written so a direct answer exists on the page.
- Do machines currently believe the right thing? Freshness, third-party mentions, a spot check of what AI assistants say today, and a dated baseline you can compare against later.
Each check has a plain-language note on how to verify it. You don’t need a crawler, an API key or a developer. You need a browser, your own website and an honest afternoon coffee.
Why does consistent naming matter so much?
Because naming is where most entity confusion starts. Consider three different eras of the same problem.
In 1753, Carl Linnaeus published Species Plantarum and popularized two-part botanical names. Before that, a single plant might carry a long descriptive phrase that changed from one author to the next. The breakthrough wasn’t better plants. It was one stable name per thing, used the same way everywhere.
Melvil Dewey’s classification system, first published in 1876, did something similar for libraries. A book’s call number didn’t depend on which librarian shelved it. Consistency made the collection navigable by strangers.
Telephone directories made the same bet for businesses. A listing was useful because the name, address and number matched what was on the shopfront. A business listed under three spellings was, functionally, three half-businesses.
Your business is in the same position with AI systems. “Smith & Co Plumbing,” “Smith and Company,” and “SmithCo” on three different profiles look to you like one company with casual branding. To a system reconciling sources, they can look like a puzzle. The naming check asks you to pick one canonical name and audit the places it appears. It’s the least glamorous fix in AEO, and frequently the most effective.
How do you read your About page like a machine would?
Open your About page and try to find, within the first two paragraphs, plain answers to four questions: who you are, what you do, who you do it for, and where you operate. If any of those answers is buried under a founder origin story, a mission statement or a carousel, a machine has to work harder to extract it. Models are good at extraction. They’re better at it when you don’t make them dig.
The Quick Check treats a clear About page as a check in its own right because it’s the page most likely to be read as your self-description. Treat it as reference material, not as a brochure. You can keep the origin story. Just put the facts first.
What does Organization schema with sameAs actually do?
Organization schema is structured data, usually JSON-LD, that states machine-readably who your business is. The sameAs property lists your official profiles: your LinkedIn page, your Google Business Profile, your Crunchbase entry, your social accounts. It’s a way of saying, in a format systems are built to read, “these are all me.”
The check isn’t “do you have schema.” It’s “does your schema exist, and does it point to the right places?” Stale sameAs links to abandoned profiles are surprisingly common. So are schema blocks copied from a template that still name the template’s sample business. (Nobody admits to this. Everybody has seen it.)
Which crawlers should you allow?
Crawler access is the check most likely to produce a surprise. A robots.txt file written years ago, or a security plugin with aggressive defaults, can block bots you actually want. The names of AI and search crawlers also change, and some control different things: for example, Google’s documentation explains that its Google-Extended token governs certain AI training uses and doesn’t control inclusion in Google Search.
The Quick Check doesn’t tell you what to allow. That’s a business decision. It tells you to look, record what you find and verify crawler names against vendor documentation on the day you check. The goal is that your crawler policy is a choice, not an accident.
How do you spot-check what AI assistants say about you?
Ask two or three assistants the questions a prospective customer would ask: what does this business do, where is it located, who runs it, what does it charge, how does it compare. Record the answers verbatim with the date. Don’t argue with the model. Just document.
What you’re looking for isn’t perfection. You’re looking for patterns:
- Absence: the assistant doesn’t know you exist.
- Confusion: it merges you with a similarly named entity.
- Staleness: it describes something you used to do.
Each pattern points to a different fix. Absence usually points to authority and third-party mentions. Confusion points to naming and sameAs. Staleness points to freshness and to the pages that still describe the old version of you.
Why does a dated baseline matter more than a score?
Because without one, every future improvement is an anecdote. The final check in the Quick Check is to record your results with a date. It’s the difference between “I think AI describes us better now” and “On this date, two of three assistants misstated our service area; on this later date, one did.”
That baseline also protects you from the most common AEO mistake: changing six things at once and having no idea which one mattered.
How do you score it, and what does the score mean?
Each check is pass or fail, and the total gives you a rough reading. The Quick Check suggests a simple threshold: if you score under 7 of 10, the foundation needs work before content strategy will pay off. That’s not a grade on your business. It’s a reading on how legible your business is to systems that are increasingly doing the first round of research for your customers.
A low score is useful news. It means the highest-leverage fixes are cheap ones: names, an About page, a schema block, a robots.txt line. Those are afternoon jobs, not quarterly initiatives.
What should you do after the test?
Here are six common next moves, in the order most businesses should consider them:
- Fix naming first. Pick the canonical name and update the profiles you control.
- Rewrite the top of your About page so the four core facts appear in plain sentences.
- Add or repair Organization schema and make sure every sameAs link resolves to a live, official profile.
- Review crawler access and turn accidental blocks into deliberate choices.
- Re-run the AI spot check after changes have had time to propagate, and compare against your dated baseline.
- Decide whether you need the full picture. If the quick test raised more questions than it answered, a structured audit is the natural next step.
That last point is where the AI-Visibility (AEO) Audit comes in. It expands the ten quick checks into 60 scored checks across eight audit areas, with a workbook that turns your scores into a prioritized fix plan. The Quick Check tells you whether there’s a problem. The Audit tells you exactly where, and what to do about it first.
Who should run this test?
Three kinds of people get the most out of it:
- Owners of small businesses who keep hearing about AI search and want a concrete first look.
- Freelance SEOs and marketers who want a fast, credible discovery step before scoping paid work.
- Agency teams who need a one-page artifact to start a conversation with a new client.
For consultants in particular, the Quick Check makes a natural opening exercise: run it together on a discovery call, and the client sees their own gaps instead of hearing about them secondhand.
The cheapest insight in AEO
Answer engine optimization has a reputation for being complicated, and parts of it are. But the foundation is mostly bookkeeping: the same name everywhere, a self-description a machine can parse, a structured statement of identity, open doors for the crawlers you want and a record of where you started.
Ten minutes is enough to find out whether yours is in order. Download the Free AI-Visibility Quick Check, print it, run it on your own site this week, and file the results with today’s date. Future you will want the receipt.
The Quick Check and the Audit both live on our Content & SEO Workflows page, with the rest of the tools in this line.
Leave a Reply