[becitable]
[On-Site Optimisation]

Entity optimisation for B2B SaaS: how AI learns who you are

Category: On-Site Optimisation · Reading time: 9 min

A single terracotta pot in soft window light, representing a clearly defined entity

AI tools don't rank your pages. They build a model of your company as an entity, and they recommend or quote you when that entity matches the user's query. Entity optimisation is the work of making that model accurate and unambiguous.

What is an entity in GEO?

An entity is a "thing" the AI has a stable identity for - a company, a product, a person, a place. For B2B SaaS, the entities that matter are your company, your product, your founders, and your category. AI tools are very good at recognising entities once they have enough corroborating signal. They're very bad at recognising them when signals are sparse or conflicting.

Why does entity optimisation matter more than keyword optimisation?

Traditional SEO optimised pages for keyword phrases. AI tools don't think in keywords - they think in entities and relationships ("Acme makes product analytics for B2B SaaS"). If the AI doesn't have a clean entity record for you, no amount of keyword optimisation will get you recommended.

Which signals does AI use to build your entity?

Entity recognition is built from corroboration across sources. AI tools cross-reference your site, your schema, your third-party profiles and press mentions. The more these agree, the stronger the entity record.

Facet
Signal AI needs to see
Name
Consistent product and company name across site, schema and external profiles.
Category
A named, established category on homepage and Organization schema.
Audience
Explicit buyer description ('for B2B revenue ops teams').
Founders / people
Real named people on About, LinkedIn, and Person schema.
Verified profiles
sameAs links to LinkedIn, G2, Capterra, X, GitHub, Crunchbase.
Reviews and ratings
Active, recent reviews on G2 or Capterra with consistent product name.
Press and mentions
Third-party articles, listicles, podcasts that name you correctly.
Wikipedia / Wikidata
Where eligible. Aspirational for most SaaS but enormous if achieved.

Why is consistency the highest-leverage move?

Decide once on:

  • Your exact product name (capitalisation, spacing, suffix).
  • Your exact company legal name vs trading name.
  • Your one-line category description.
  • Your one-line audience description.

Then propagate them everywhere: homepage, About, schema, LinkedIn, G2, Capterra, Crunchbase, press kit. AI tools treat inconsistency as ambiguity and downgrade you accordingly.

Why sameAs is the most important schema property

The sameAs array in your Organization schema is the single most powerful entity signal you control. It tells AI tools "this entity on my site is the same entity as this LinkedIn profile, this G2 page, this Crunchbase record." It collapses ambiguity at a stroke.

Include at least: LinkedIn company page, X profile, G2 listing, Capterra listing, GitHub org, Crunchbase record. If you have a Wikipedia entry, link to that too.

What if my product name is ambiguous?

If your product shares a name with a common word ("Notion", "Linear", "Loop", "Tally") or with another SaaS, you have an entity disambiguation problem. The fixes:

  • Always pair the product name with the company name in the first mention of any page.
  • Strengthen sameAs as much as possible.
  • Use a distinctive alternateName in schema (e.g. "Acme - revenue ops platform").
  • Prioritise getting on G2 and Capterra. They disambiguate aggressively.

How do founders and team fit into your entity?

AI tools weight named, verifiable people. Your About page should name the founders, link to their LinkedIn profiles, and ship Person schema with title, employer and sameAs. "Founded by ex-Google engineers" is filler. "Founded by Jane Smith (ex-Google, ex-Stripe)" is entity signal.

An entity audit you can run in 30 minutes

  1. Ask ChatGPT and Perplexity: "what is [product name]?" Compare to your homepage description.
  2. Ask: "who founded [product]?" and "what category is [product] in?" Note any inaccuracies.
  3. Audit your sameAs array. Add any missing profile (LinkedIn, G2, Capterra, Crunchbase).
  4. Check that product name, category and audience are identical across homepage, About, schema and LinkedIn.
  5. Fix the worst inconsistency first.

Related: schema markup that actually helps AI cite you and why ChatGPT keeps recommending your competitors.