<!-- Sources of truth:
Mailchimp: app/samples/mailchimp-lite-2026-10-15-v2.md and -v2.json (re-derived 2026-10-15T16:22:41Z from the run created 16:18:34Z = 12:18 PM ET). module4 identical to v1.
Our own site: drafted from app/samples/answercheckup-lite-2026-10-15-v2.md/.json, then re-checked against the newest app/samples/answercheckup-lite-2026-10-15-v3.md/.json (Builder Lite v3, 12:23 PM ET; re-derived from the run created 16:07:57Z = 12:07 PM ET). Score 25/100, all 11 per-check points, the 0 of 12 evidence and both quotes are identical in v1, v2 and v3; v3 only changes the other-names list, which this post does not use.
Comments that say ".md" or ".json" mean the -v2 files.
8 not found + 4 ambiguous = 12: one tag per answer, per-engine reason tags not_found 4+4 and ambiguous 2+2, confirmed by Builder 12:29 PM ET (report shows only these counts in the coming -v4 file). The "8 ... 10" entity-row figures are overlapping pattern flags and are not quoted. -->
Entity signals: why AI can describe some brands and not others
AI answer engines describe a brand clearly when many independent signals agree on what that brand is: its own site's structured data, an About page, a Wikipedia article, consistent naming, and other sites that talk about it. When those signals are thin, the engines hedge, ask which one you mean, or say they can't find you.
We scored two sites on the same 11-point checklist on Oct 15, 2026. Mailchimp scored 70 out of 100. Our own site, answercheckup.com, scanned with our own tool on the same day, scored 25 out of 100. Here is what made the difference.
<!-- source: mailchimp v2 .json module4.score = 70, maxScore 100, 11 checks; answercheckup v2/v3 .json module4.score = 25, 11 checks -->
About this example. The Mailchimp results are an independent sample we made with public API answers on Oct 15, 2026. Answer Checkup is not affiliated with Mailchimp or with any other brand named in this post, and none of them asked for or reviewed it.
What "entity" means here
An entity is a thing an engine can identify without guessing: this name, this company, this website. Search engines and answer engines try to connect a name to one specific business. Signals that help:
- Structured data on your homepage: schema.org Organization with your name, your URL, your logo, contact details and sameAs links to your other profiles.
- Basic pages: an About page that says who you are and a Contact page.
- One spelling of your name across your title, your schema and your llms.txt.
- Reference sources like Wikipedia and Wikidata, where they genuinely apply.
- Answers themselves: when asked by name, can the engines describe you?
None of these is a switch that makes an engine recommend you. Together they make it easier for an engine to be sure who you are.
The two checklists side by side
| Check | Mailchimp | Our own site |
|---|---|---|
| Organization schema on the homepage | 15/15 | 15/15 |
| Organization has an absolute url on this domain | 5/5 | 0/5 |
| Organization has a logo | 0/5 | 0/5 |
| Organization sameAs links to other profiles | 0/10 | 0/10 |
| Organization has contact data | 0/5 | 0/5 |
| Brand name written the same way across title, schema, llms.txt | 5/10 | 5/10 |
| About page | 10/10 | 0/10 |
| Contact page or visible contact email | 10/10 | 5/10 |
| Wikipedia article with the brand name | 10/10 | 0/10 |
| Wikidata item with the brand name | 5/10 | 0/10 |
| Engines can describe the brand when asked by name | 10/10 | 0/10 |
| Total | 70/100 | 25/100 |
<!-- source: mailchimp v2 .md "Module 4" table, points column; answercheckup v2/v3 .md "Module 4" table, points column; both .json module4.checks[].points. Sums: 15+5+0+0+0+5+10+10+10+5+10 = 70; 15+0+0+0+0+5+0+5+0+0+0 = 25 -->
A score is our own simple checklist, partly heuristic, not an industry standard. A low score means fewer clear signals, not a bad business.
What the engines said when asked by name
We asked each engine 6 questions that use the brand name, like "What is [brand]?" and "Is [brand] legit?", for 12 answers per brand.
For Mailchimp, 12 of 12 answers named and described it. One answer opened: "Mailchimp is a popular all-in-one marketing platform best known for email marketing."
<!-- source: mailchimp v2 .md check "Engines can describe the brand when asked by name": "12 of 12 branded answers name and describe the brand"; quote from .md "How the engines describe the brand" (prompt: What is Mailchimp?). Punctuation shown in ASCII as in the .md. -->
For our own site, 0 of 12 answers named and described it. 8 couldn't find us, and 4 treated the name as ambiguous. Typical replies:
- "I couldn't find a well-known product called exactly "Answer Checkup.""
- ""Answer Checkup" is ambiguous, and the search results don't identify a specific product or service by that exact name."
<!-- source: answercheckup v2/v3 .md check engines_know_brand "0 of 12 branded answers name and describe the brand"; quotes from v2 .md "How the engines describe the brand" (prompts: What are the best alternatives to Answer Checkup?; What is Answer Checkup?). Curly quotes normalized to ASCII. -->
That's not a surprise for a new site. But it shows the gap clearly: the engines had nothing solid to tie our name to.
What separates 70 from 25
Outside sources. Mailchimp has an exact-title Wikipedia article and a Wikidata item with the same label (our check didn't confirm the Wikidata item is the same business, so it got half points). We have neither. Those come from being notable and written about, not from anything you can add to your own site. Don't create them for yourself; Wikipedia has its own rules on notability and conflicts of interest.
<!-- source: mailchimp v2 .md wikipedia pass "exact-title article found", wikidata warn 5/10 "not verified to be this business"; answercheckup v2/v3 .md wikipedia fail, wikidata fail -->
Basic pages. Mailchimp has an About page and a Contact page. Our site had no About page linked from the homepage, and no Contact page, though an email address was listed. That's 15 points we can fix in an afternoon.
<!-- source: mailchimp v2 .md about_page pass, contact pass; answercheckup v2/v3 .md about_page fail "none linked from the homepage; probed https://answercheckup.com/about (404)" (URL not published here), contact warn 5/10 "no contact page, but an email address is on the homepage or in llms.txt". 10 + 5 = 15 points. -->
Engines recognizing the name. This is the result of the other signals, not a separate fix. A clear, well-linked entity is easier to describe.
What a well-known brand still misses
Mailchimp's homepage Organization schema had no logo, no sameAs links and no contact data. Those three checks are 20 of the 30 points it missed. Its name was also missing from the homepage h1, which cost 5 more.
<!-- source: mailchimp v2 .md: org_logo 0/5, org_sameas 0/10, org_contact 0/5 = 20 of 30 missed; name_consistency warn 5/10 "h1: no"; wikidata 5/10 is the remaining 5. 100 - 70 = 30. Next step in v2 .json nextSteps[0]: "Add a logo, sameAs links to your real profiles and contact data ... to the homepage Organization JSON-LD." -->
So a brand can be famous and still leave easy, on-site signals blank. Fame covers for it. A small business doesn't have that cover, which is why the on-site signals matter more for you.
What to fix first
These are the on-site signals, in the order we'd do them. They're also what we're fixing on our own site.
- Complete your homepage Organization schema. Name, absolute URL, logo, contact data, and an absolute @id. See schema.org/Organization.
- Add sameAs links to real profiles you own, like a LinkedIn company page or GitHub organization, where they fit your business.
- Publish an About page that says who runs the business, what it does, and how to reach you, and link it from the homepage.
- Publish a Contact page.
- Write your name one way in your title, your og:site_name, your schema, your llms.txt and your homepage h1. Our llms.txt guide covers that file.
- Earn outside mentions honestly. Real listings, real reviews, real coverage. The AI SEO guide has more on this.
<!-- source for item list: mailchimp v2 .json nextSteps and answercheckup v2/v3 .json nextSteps (org_schema, org_sameas, about_page, contact); og:site_name gap from answercheckup v2/v3 .md name_consistency "Missing sources: og:site_name" -->
To check whether engines mention you at all, start with does ChatGPT recommend my business?
Limits of this comparison
- One run is a snapshot. Each branded question was asked once per engine.
- The API is not the app. We measured through the OpenAI API with web search and the Perplexity API, which approximate the consumer apps but are not identical.
- Some checks are heuristic. Name consistency and "engines can describe the brand" are pattern-based. The Wikidata check matches a label and doesn't confirm it's the same business.
- Not checked: Google's Knowledge Graph, third-party listings such as review sites and company directories, and the consumer apps themselves.
- Unequal brands. One is a long-established company, the other a new site. The comparison shows which signals differ, not who is better.
<!-- source: both v2 .md "Not checked: Google Knowledge Graph ...; Third-party listings ...; Consumer ChatGPT / Perplexity apps ...; Whether a Wikidata item with the same label is this business" -->
FAQ
- What are entity signals?
- The clues that tell an engine which specific business a name refers to: structured data, About and Contact pages, consistent naming, reference sources and outside mentions.
- Why can't ChatGPT describe my business?
- Often because there are too few signals tying your name to your site, or because your name is shared with something else. Check your homepage schema, your About page and how consistently your name is written.
- Does Organization schema make AI recommend me?
- No. It helps engines connect your name to your site. Whether they recommend you depends on much more, including what other sites say about you.
- Do I need a Wikipedia article?
- No, and you shouldn't write one about yourself. It helps well-known brands, but most small businesses won't qualify. Focus on the signals you control.
- What is sameAs?
- A schema.org property that lists URLs of your other official profiles, so an engine can confirm they belong to the same business.
Sources
- Answer Checkup Lite sample for mailchimp.com, run Oct 15, 2026 at 12:18 PM ET, through the OpenAI API with web search and the Perplexity API. An independent sample, not affiliated with or reviewed by Mailchimp.
- Answer Checkup Lite report for our own site, answercheckup.com, scanned with our own tool on Oct 15, 2026 at 12:07 PM ET.
- schema.org Organization: https://schema.org/Organization
- schema.org sameAs: https://schema.org/sameAs
Check your own site
Answer Checkup is not affiliated with OpenAI, Perplexity, Google, Mailchimp, Wikipedia, Wikidata, LinkedIn or GitHub.
Questions: [email protected]