Checklist · AI search optimization (SEO+)
Entity and brand consistency checklist
A checklist for giving a company and its people one name and one set of facts across their site, markup, profiles, and the sources AI assistants read.
By Paul Maxwell, founder of AEO HQ
Published · Updated
An entity and brand consistency checklist tests whether a company and the people behind it appear under one name, with one set of facts, everywhere that search engines and AI assistants read about them: the company's own site and markup, the profiles it controls, and the sources others maintain. This one has 14 checks in six groups. The problem it guards against is documented: language models often "incorrectly merge information belonging to different entities" (opens in a new tab) that share a name. No study we found tests whether consistency work changes AI recommendations, so each check states how strong its evidence is.
An entity is a distinct thing, such as a company, a person, or a product, that search systems try to tell apart from others. This checklist belongs to AEO HQ's guide to AI search optimization. It expands the entity group of the AEO checklist, whose two checks cover one name and description everywhere (check 5.1) and Organization markup on the home page (check 5.2), and it does not repeat them. The sources were read on September 27, 2026.
Scope
- Covers: the names, descriptions, and core facts of a company, its products, and its founders and authors, on the company's site, in its structured data, on the profiles it controls, and in sources others maintain, such as Google's knowledge panels and Wikipedia.
- Does not cover: earning mentions and reviews, which are covered in brand mentions and AI recommendations and in the AEO checklist's third-party checks; crawler access and indexing, in the technical SEO checklist for AI search; and page content, in the GEO checklist.
- For: founders and marketers at B2B companies, especially those whose company or founder shares a name with someone else.
- Pass criteria are AEO HQ's recommendations. The facts behind them link to their sources.
- Evidence labels:
- Strong: official platform documentation, or several independent studies that agree.
- Moderate: laboratory studies, a single study, or correlations.
- Weak: one small test, or a source with a conflict of interest.
- What the evidence covers: the evidence that assistants confuse or misdescribe entities is moderate. For the fixes, it is mostly platform documentation: in AEO HQ's research review, we found no study that tests whether entity markup, profile links, or consistent descriptions change which companies AI assistants recommend.
Prerequisites
- An entity record: one document that lists the company's official name; other names it goes by, such as abbreviations and former names; its legal name; a one-line description; the canonical domain; the founding date; founders and key people with their titles; the address or service area; contact details; company identifiers such as registration, VAT, DUNS, or LEI numbers, where they exist; the official profile URLs; and the logo file. It extends the fact sheet in the AEO checklist's prerequisites.
- Admin access to the site's code or content system, and to each official profile.
- A list of branded questions to ask AI assistants (check 6.2).
Quick checks to run first
If time is short, AEO HQ recommends starting with these five checks:
- Check 1.1: other entities that share the name are known.
- Check 1.2: Google shows the official name as the site name.
- Check 2.2: Organization markup carries identifying details.
- Check 3.1: the official profiles and the site link to each other.
- Check 6.2: assistants are tested for merges and wrong facts.
1. The name
- 1.1 Other entities that share the name are known and recorded. Verify: search the company name and each founder's name in Google and Bing, and ask three AI assistants "What is [company]?" and "Who is [founder]?". Record every other company, product, person, or abbreviation with the same or a similar name. Pass: the entity record lists each one, with the distinguishing detail to use beside the name, such as the domain, the category, or a location; the AEO checklist's check 5.1 then requires that detail near the name. Why: in a benchmark built from Wikipedia's disambiguation pages, models "often yield ambiguous answers or incorrectly merge information belonging to different entities" (opens in a new tab) (peer-reviewed), and in a test of 16 language models, their grasp of facts was "far from being perfect," "especially for facts of torso-to-tail entities," (opens in a new tab) the less well-known ones (peer-reviewed). For example, when AEO HQ searched for its own name on September 27, 2026, the results were pages about American Eagle Outfitters and a disambiguation page for "AEO," with no result for aeohq.ai. Evidence: moderate that confusion happens; weak for the effect of a descriptor.
- 1.2 Google shows the official name as the site name. Verify: search Google for the domain and read the site name shown above the result. Then view the home page's source for WebSite structured data, the
og:site_nametag, the<title>, and the H1. Pass: Google shows the official name; the home page has WebSite markup withnameandurl, plusalternateNamefor any common abbreviation; and theog:site_nametag, the title, and the headings use the same name. Why: Google generates site names automatically from "content from a site's home page and references to it that appear on the web" (opens in a new tab). It says the WebSite structured data "must be on the home page of the site" (opens in a new tab), asks owners to "Use your site name consistently across your home page" (opens in a new tab) and to choose "a unique name that accurately reflects the identity of your site and isn't misleading for users" (opens in a new tab), and says analternateNamelets Google "consider other options if your preferred choice isn't selected" (opens in a new tab). Evidence: strong (official documentation); the effect on AI answers has not been measured. - 1.3 Former names and domains lead to the current ones. Verify: list former company names, product names, and domains, request pages on each old domain, and search each old name. Pass: old pages return a permanent redirect to the matching current page; profiles and listings show the current name; and the Organization markup's
alternateNamelists any former name that people still use. Why: with a permanent redirect, Google shows "the new redirect target in search results" (opens in a new tab), and Bing asks for 301 redirects for permanent URL changes (opens in a new tab). Google's Organization markup takesalternateNamefor "Another common name that your organization goes by, if applicable" (opens in a new tab). Evidence: strong for the redirect guidance; weak for any effect on AI answers.
2. The company's own site
- 2.1 One page states who the company is, in text. Verify: open the home page or About page with scripts turned off. Pass: it states, in visible text, the official name, what the company sells and to whom, where it operates, who founded or runs it, and how to contact it. Why: Bing asks site owners to "Use clear and consistent naming for people, organizations, products, and locations" (opens in a new tab) and to make sure "Important information is visible on the URL itself" (opens in a new tab). Google recommends placing organization information "on your home page, or a single page that describes your organization, for example the about us page" (opens in a new tab). Evidence: strong (official documentation).
- 2.2 Organization markup carries identifying details, not only a name and logo. Verify: read the JSON-LD on that page, and test it with Google's Rich Results Test. Pass: it uses the most specific Organization subtype that fits;
nameandalternateNamematch the site name;legalNameappears if it differs fromname;urlis the canonical home page;logois at least 112 by 112 pixels;addressortelephoneappears where one exists; identifiers such asvatID,taxID, oriso6523Codeappear where they exist; and every value is also visible on the page (the AEO checklist's check 6.1). The sameAs list is check 3.1. Why: Google recommends "the most specific schema.org subtype" of Organization that matches the organization, and a logo of at least 112 by 112 pixels at a crawlable, indexable URL (opens in a new tab). It says some Organization properties are "used behind the scenes to disambiguate your organization from other organizations (like iso6523 and naics)" (opens in a new tab). It recommends "name or alternateName for your business name as well as an indication of real-world presence (for example, address or telephone) and online presence (for example, url or logo)" (opens in a new tab), saysurl"helps Google uniquely identify your organization" (opens in a new tab), and asks sites to "Use the same name and alternateName that you're using for your site name" (opens in a new tab). Markup alone did not raise AI citations in the one matched study: after 1,885 pages added JSON-LD markup, their AI Overview citations fell 4.6% against 4,000 matched control pages, and changes in AI Mode and ChatGPT were statistically indistinguishable from zero (opens in a new tab) (vendor study; August 2025 to March 2026). Evidence: strong for Google's documentation; weak for AI answers.
3. Profiles the company controls
The AEO checklist's check 5.1 compares names and descriptions across profiles. The checks below cover the rules that platforms document.
- 3.1 The official profiles and the site link to each other. Verify: compare the
sameAsvalues in the Organization markup, and in each founder's Person markup, with the list of official profiles in the entity record. Then open each profile and read its website field. Pass: every official profile is listed, no unofficial or dead URL is listed, and each profile links to the canonical domain. Why:sameAsis for "The URL of a page on another website with additional information about your organization," (opens in a new tab) Google says, "for example, a URL to your organization's profile page on a social media or review site" (opens in a new tab). The link back from each profile is AEO HQ's recommendation: a link in both directions lets a reader, or a program, confirm that both pages describe the same company. Evidence: strong for Google's documentation; we found no study of the effect of links in both directions. - 3.2 The Google Business Profile, if the business is eligible, follows Google's rules. Verify: check eligibility first. A business can have a Business Profile only if it "either has a physical location that customers can visit, or travels to customers where they are" (opens in a new tab); an online-only company that neither receives customers nor travels to them does not qualify, and can skip this check. Then open the profile. Pass: there is one profile for the business, or one for each eligible location; the name is the real-world name, with no taglines, service terms, or locations added; the category is specific; the address or service area is accurate; and the profile is verified. Why: Google asks businesses to "Represent your business as it's consistently represented and recognized in the real world across signage, stationery, and other branding" (opens in a new tab), says "There should only be one profile per business" (opens in a new tab), and says the name "should reflect your business's real-world name, as used consistently on your storefront, website, stationery, and as known to customers" (opens in a new tab). Google says Business Profiles can help "products and services to be visible in both AI responses and other Google Search results" (opens in a new tab), and it lists "Checking that your Merchant Center and Business Profile information is up-to-date" (opens in a new tab) among the SEO basics for its AI features. Google says verifying a profile makes the business "more likely to show up in search results" (opens in a new tab). Evidence: strong (official documentation).
- 3.3 A knowledge panel, if Google shows one, is claimed and correct. Verify: search Google for the company name and each founder's name, and look for a knowledge panel, the information box Google shows about some people and organizations. Pass: if a panel exists, it has been claimed, and any error has been reported through its feedback option, with a source. Why: Google's steps are to "Search for yourself or the entity you represent and find your knowledge panel" (opens in a new tab), then "At the bottom, click Claim this knowledge panel" (opens in a new tab), and to sign in to one of the official sites or profiles it lists, such as YouTube or Search Console. Google says panel information "generates automatically based on public information on the web" (opens in a new tab), that its "current policy doesn't manually create or delete Knowledge Panels" (opens in a new tab), and that for a verified user whose panel is about them, "we can prioritize your feedback" (opens in a new tab). Organization markup can influence which logo appears in the knowledge panel (opens in a new tab). Evidence: strong (official documentation); the effect on AI answers has not been measured.
4. Sources others maintain
What reviewers, directories, and the press say about a company is covered by the AEO checklist's third-party checks. The checks below cover how the company handles those sources.
- 4.1 The company does not write or edit its own Wikipedia article. Verify: check whether Wikipedia has an article about the company or its founders, and ask whether anyone at the company, or paid by it, has edited it. Pass: no one at the company or paid by it edits the article directly, and corrections are proposed on the article's talk page, with sources and the connection disclosed. Why: Wikipedia's conflict-of-interest guideline says: "Do not edit Wikipedia in your own interests, or in the interests of your external relationships. Paid editing must be disclosed." (opens in a new tab) It says editors with a conflict of interest are "strongly discouraged from editing affected articles directly" (opens in a new tab) and "may propose changes on talk pages" (opens in a new tab). We found no study that tests whether a Wikipedia article or a Wikidata entry changes AI recommendations. Evidence: strong for Wikipedia's rules; none for the effect on AI answers.
- 4.2 Press releases, partner pages, and outside bios state current facts. Verify: collect the past year's press releases, partner and marketplace listings, podcast and event pages, and author bios on other sites. Pass: each gives the current name, description, founder names and titles, and website, and none states an old price, product, or location; outdated ones have a correction requested. Why: Bing asks site owners to "Remove or revise outdated information to prevent incorrect information from surfacing" (opens in a new tab). What a model learns in training also depends on how facts are repeated: in controlled experiments, a fact could be reliably extracted only if it was "sufficiently augmented (e.g., through paraphrasing, sentence shuffling) during pretraining" (opens in a new tab) (peer-reviewed; laboratory). Our reading is that the facts should match everywhere, stated in each source's own words, rather than one copied paragraph. That reading has not been tested on companies. Evidence: strong for Bing's guidance; weak for the effect on AI answers.
5. People
- 5.1 Each founder and author has one profile page, and every byline points to it. Verify: follow the byline on several pages, and read the Article and profile page markup. The AEO checklist's check 4.3 covers the profile page itself. Pass: every byline for a person links to the same profile page; the Article markup lists every author shown on the page, with
author.nameholding only the person's name; and the profile page carries ProfilePage markup whose Person hassameAslinks to the person's other profiles. Why: Google asks site owners to "Make sure that all the authors that are presented as authors on the web page are also included in markup" (opens in a new tab), to put only the author's name inauthor.name, not the publisher's name or the author's job title (opens in a new tab), and, when the author URL is an internal profile page, to mark that author up "using profile page structured data" (opens in a new tab). It says "Google can understand both sameAs and url when disambiguating authors" (opens in a new tab). Evidence: strong for Google's documentation; weak for AI answers. - 5.2 Each person uses one form of their name and one title everywhere. Verify: compare the person's name, title, and company on the company site, their professional profiles, conference and podcast pages, and author bios on other sites. Pass: the same form of the name (for example, with or without a middle initial), the same current title, and the same company name appear everywhere, and bylines and bios put the company's name next to the person's. Why: a person who shares a name with others is exposed to the merging described in check 1.1, and AI-written descriptions of people have been inaccurate: in a 2023 evaluation, only 58% of the atomic facts in ChatGPT's biographies of people were supported by a reliable knowledge source (opens in a new tab) (peer-reviewed; the models tested have since changed). Evidence: moderate that people are misdescribed; weak for the effect of consistency.
6. Upkeep and testing
- 6.1 One person owns the entity record, and every change reaches every surface. Verify: take the last three changes to a core fact, such as a price, an address, a product name, or a leader, and check each surface: the site, the markup, the profiles, the Business Profile, directories, and the llms.txt file if the site has one. Pass: the record names an owner, each change is dated in it, and every surface shows the new fact. Changed pages are sent through IndexNow, the AEO checklist's check 1.5. Why: Bing asks site owners to "Update content when facts or guidance change" (opens in a new tab) and to "Notify IndexNow when content is added, updated, or deleted" (opens in a new tab). Google says that if business information "isn't accurate, your Business Profile might not show up for relevant searches in your area" (opens in a new tab). Evidence: strong for the platform guidance; no study has measured how quickly AI answers pick up a change.
- 6.2 Assistants are tested for merges and wrong facts about the company and its people. Verify: each month, ask each assistant a fixed set of branded questions several times, such as "What is [company]?", "Who founded [company]?", "How much does [company] cost?", and "Who is [founder]?". Ask each once with the name alone and once with the distinguishing detail from check 1.1. Pass: every answer that mixes in another entity's facts, or states a wrong fact, is logged with the source it cited and traced as in the GEO checklist's accuracy check. Grading answers against the fact sheet is the AEO checklist's check 9.4. Why: merging entities that share a name is a documented failure (check 1.1), and answers change "across runs, prompts, and time, making one-off observations unreliable" (opens in a new tab) (preprint), so one test is one sample. How to measure AI visibility describes a full design. Evidence: moderate.
Version and change log
- Version 1.0, September 28, 2026: First published. Sources checked on September 27, 2026.
Next steps
AEO HQ's Instant AEO Audit ($499) covers a small part of this checklist: it records which schema.org types the home page's JSON-LD declares, such as Organization, but it does not validate the markup or compare facts across profiles. Its limits are listed in our methodology.
Sources
- Lee, Y., Ye, X., & Choi, E. (2024). AmbigDocs: Reasoning across documents on different entities under the same name. In Proceedings of the First Conference on Language Modeling (COLM 2024). https://doi.org/10.48550/arXiv.2404.12447 (opens in a new tab)
- Sun, K., Xu, Y., Zha, H., Liu, Y., & Dong, X. L. (2024). Head-to-Tail: How knowledgeable are large language models (LLMs)? A.K.A. will LLMs replace knowledge graphs? In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) (pp. 311–325). Association for Computational Linguistics. https://doi.org/10.18653/v1/2024.naacl-long.18 (opens in a new tab)
- Google. (2025, December 10). Provide a site name to Google Search. Google Search Central. https://developers.google.com/search/docs/appearance/site-names (opens in a new tab)
- Google. (2026, April 14). Redirects and Google Search. Google Search Central. https://developers.google.com/search/docs/crawling-indexing/301-redirects (opens in a new tab)
- Microsoft Bing. (n.d.). Bing Webmaster Guidelines. Retrieved September 27, 2026, from https://www.bing.com/webmasters/help/webmaster-guidelines-30fba23a (opens in a new tab)
- Google. (2026, September 8). Organization schema markup. Google Search Central. https://developers.google.com/search/docs/appearance/structured-data/organization (opens in a new tab)
- Linehan, L. (2026, May 11). We tracked 1,885 pages adding schema. AI citations barely moved. Ahrefs. https://ahrefs.com/blog/schema-ai-citations/ (opens in a new tab)
- Google. (n.d.). Guidelines for representing your business on Google [Google Business Profile Help]. Retrieved September 27, 2026, from https://support.google.com/business/answer/3038177 (opens in a new tab)
- Google. (2026, July 10). Optimizing your website for generative AI features on Google Search. Google Search Central. https://developers.google.com/search/docs/fundamentals/ai-optimization-guide (opens in a new tab)
- Google. (2025, December 10). AI features and your website. Google Search Central. https://developers.google.com/search/docs/appearance/ai-features (opens in a new tab)
- Google. (n.d.). Tips to improve your local ranking on Google [Google Business Profile Help]. Retrieved September 27, 2026, from https://support.google.com/business/answer/7091 (opens in a new tab)
- Google. (n.d.). Get verified on Google [Knowledge Panel Help]. Retrieved September 27, 2026, from https://support.google.com/knowledgepanel/answer/7534902 (opens in a new tab)
- Google. (n.d.). Submit feedback on content about you [Knowledge Panel Help]. Retrieved September 27, 2026, from https://support.google.com/knowledgepanel/answer/7534842 (opens in a new tab)
- Wikipedia:Conflict of interest. (n.d.). In Wikipedia. Retrieved September 27, 2026, from https://en.wikipedia.org/wiki/Wikipedia:Conflict_of_interest (opens in a new tab)
- Allen-Zhu, Z., & Li, Y. (2024). Physics of language models: Part 3.1, knowledge storage and extraction. In Proceedings of the 41st International Conference on Machine Learning (Proceedings of Machine Learning Research, Vol. 235, pp. 1067–1077). https://proceedings.mlr.press/v235/allen-zhu24a.html (opens in a new tab)
- Google. (2026, September 8). Learn about article schema markup. Google Search Central. https://developers.google.com/search/docs/appearance/structured-data/article (opens in a new tab)
- Min, S., Krishna, K., Lyu, X., Lewis, M., Yih, W.-t., Koh, P. W., Iyyer, M., Zettlemoyer, L., & Hajishirzi, H. (2023). FActScore: Fine-grained atomic evaluation of factual precision in long form text generation. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics. https://arxiv.org/abs/2305.14251 (opens in a new tab)
- Schulte, J., Bleeker, M., & Kaufmann, P. (2026). Don't measure once: Measuring visibility in AI search (GEO) (arXiv:2604.07585) [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2604.07585 (opens in a new tab)
How to cite this page
Maxwell, P. (2026). Entity and brand consistency checklist. AEO HQ. Last updated September 28, 2026. https://www.aeohq.ai/articles/entity-checklist
More in AI search optimization (SEO+)
Complete guide
AI search optimization: SEO plus corroboration, facts, and measurement
What AI search optimization is, how AI assistants choose sources, which tactics hold up in research, and how to measure results. Every claim is sourced.
Guide
Brand mentions and AI recommendations
What brand mentions are, where AI assistants find them, what research shows about mentions and AI recommendations, and how to earn and track them.
Guide
How AI agents find and buy services
How AI agents research and buy services in 2026: what ChatGPT, Google, Copilot, Perplexity, Claude, and Stripe support, and what a service business can do.
Guide
llms.txt: what it is and whether it matters
What llms.txt is, who reads it, and whether it helps AI visibility: the proposal, Google's position, server-log studies, and when a file is worth keeping.
Checklist
Technical SEO checklist for AI search
A technical SEO checklist for AI search: robots.txt, status codes, canonical URLs, redirects, rendering, page elements, and crawler identity, with sources.
Antipatterns
SEO antipatterns
Twelve SEO mistakes that also keep pages out of AI answers, from blocked crawlers to facts hidden in JavaScript, with the evidence and a test for each.