Entity

What is an entity?

An entity is anything search engines and AI tools understand as one distinct thing, with facts attached to it. A company, a product, a person, a city and an idea can all be entities. Google introduced its Knowledge Graph in 2012 to store them, and it described the shift as "things, not strings".

The difference matters because words are messy. When someone types "linear", they might mean the math term or Linear, the project tracking tool for software teams. Google and AI tools look at the rest of the search and what they know about each entity to decide which one the person means.

Here's what that looks like for a SaaS company. Let's say you run Acme, a hypothetical invoicing tool for creative agencies. Google and ChatGPT need to know that Acme is a company that makes invoicing and time tracking software. They also need to know it serves agencies with 10 to 200 people and is based in Austin. Then they can connect you to the right searches. If they don't, a search for "agency invoicing software" may never lead to your name.

Why entities matter for B2B SaaS

AI answers are built from what models know about entities. When a buyer asks Perplexity for "the best invoicing tools for agencies", the model needs to know your product exists, what category it belongs to and who it fits. If those facts are fuzzy or missing, you won't show up, however good your blog is.

Many SaaS names make this harder. Brands like Notion, Ramp, Clay and Linear share their names with everyday words, so search engines need strong signals to tell the company apart from the dictionary.

Here's why I think it's worth the work:

• It gets you into the right answers: Clear entity facts help AI tools place you in your category and recommend you for the use cases you really serve. This is a big part of GEO.
• It keeps the facts right: If your pricing, founding year or product line is described differently across the web, AI answers may repeat the wrong version.
• It helps branded search: When Google knows your company as an entity, searches for your name are more likely to show your site, your profiles and sometimes a Knowledge Panel.

How to build a clear entity

The aim is simple: say the same true things about your company everywhere it appears. Here's my process:

• Write one plain description: Agree on one or two sentences that say what you do, for whom and in what category. Use it on your homepage, your About page and in every profile.
• Add Organization schema: Put schema markup on your homepage or About page with your name, logo, URL and description. Google recommends adding sameAs links that point to your profiles on other sites.
• Fill in your profiles: Make your LinkedIn company page, G2 and Capterra listings, Crunchbase profile and GitHub organization match the same description, category and facts.
• Earn mentions on trusted sites: Get covered in industry press, listed in roundups and quoted in reports. Every brand mention that describes you the same way adds weight.
• Connect your people: Give your founders and experts author pages that link to their LinkedIn profiles, and mention them on your About page. People are entities too, and they lend weight to the company.
• Fix the wrong facts: Search your brand in Google, ChatGPT and Perplexity and note anything outdated. Update the source pages, like an old press release or a stale directory listing.

An entity example

Linear is a good real example of a brand that shares its name with a common word. Linear calls itself a product development system for planning and building products, in plain words. A clear category line like that, repeated across its profiles and the many articles that mention it, gives search engines and AI tools what they need to keep the company apart from the math term.

Here's a hypothetical entity check for Acme, the invoicing tool:

Once every row matches, AI tools have one clear story to repeat. Our GEO checklist includes the schema and entity checks we run for this.

How to measure entity strength

There's no single entity score, but you can check how well Google and AI tools understand you. Here's what I track:

• AI answer accuracy: Ask ChatGPT, Perplexity, Gemini and Claude "What is [your company]?" and "Who is [your company] for?" every month. Note any wrong category, pricing or audience.
• Category prompts: Track how often you're named for prompts like "best invoicing software for agencies". Being known but never recommended means your category signal is weak.
• Knowledge Panel: Search your brand name in Google and see whether a panel shows up and what it says. If it's wrong, you can claim it and suggest changes.
• Branded search: Watch searches for your name in Google Search Console. When more people know you, more of them search for you by name.

Common entity mistakes

• Describing yourself differently everywhere: "AI platform" on LinkedIn, "workflow tool" on G2 and "fintech" on Crunchbase leaves models guessing. Pick one description and use it everywhere.
• Using slogans instead of facts: A homepage that says "Work, reimagined" tells a model nothing. Lead with the plain category and audience, and keep the slogan for the design.
• Skipping schema: Organization markup takes an hour to add. Without it, you leave Google to work out your logo, profiles and description on its own.
• Ignoring old listings: A stale directory page with your old name or pricing can keep showing up in AI answers for months. Track down and fix the sources.
• Forgetting your people: Founders who post on LinkedIn and speak on podcasts build their own entities. Link them to the company so that trust carries over.

Here's a test you can run in two minutes. Ask ChatGPT "What does [your company] do, and who is it for?" Does the answer match the sentence on your homepage?