# The GEO Templates Pack

The working templates behind Entity Authority and Category Content — chapters 02 and 06 of the GEO Remediation Playbook (gtmoperator.org/geo-playbook). Free, by James Gilbert / GTM Operator.

How to use: fill these in once, then enforce them everywhere. Consistency is the signal AI engines read.

---

## 1. The canonical description

One sentence that defines your entity. Write it once, use it verbatim on your site, LinkedIn, directories, press boilerplate, and email signatures. Sameness across sources is what teaches engines who you are.

**Template:**
> [Company] is the [category noun] that [one-line outcome] for [specific audience].

**Rules:**
- One sentence. If it needs a second, the first isn't done.
- Name the category — engines classify before they recommend.
- No adjectives that can't be verified ("leading," "innovative" are noise).
- If you share a name with anyone, add the disambiguator below wherever space allows.

**Disambiguation snippet:**
> [Company] ([domain]) is the [category] by [founder/parent] — distinct from [the other entity] ([what it is]).

---

## 2. The entity consistency checklist

Run quarterly. Every row should show the same name, same description, same logo, same links.

| Surface | Canonical description? | Logo current? | Links to site? | Last checked |
|---|---|---|---|---|
| Website (About + footer) | | | | |
| Organization JSON-LD on homepage | | | | |
| LinkedIn company page | | | | |
| Crunchbase | | | | |
| G2 / review platforms | | | | |
| Google Business Profile (if local) | | | | |
| Press boilerplate (latest release) | | | | |
| Founder LinkedIn bio | | | | |
| Podcast/speaker one-liner | | | | |
| Wikipedia / Wikidata (only if genuinely notable) | | | | |

Generate the matching JSON-LD at gtmoperator.org/schema-generator.

---

## 3. The FAQ page skeleton

For the questions buyers actually ask AI (playbook chapter 06). One page per question cluster. Structure for retrieval: each section must stand alone.

```
# [The question, in the buyer's words]

[Direct answer in the first two sentences. Name the entities — no "it" or "they"
 standing in for the company or product. State the scope: who this applies to,
 as of when.]

## [Sub-question 1 — a follow-up a buyer would ask next]
[Self-contained answer, 50–150 words. Entity names visible. One claim per paragraph.]

## [Sub-question 2]
[Same rules.]

## What the critics say
[The honest counterargument and where it's right. Engines summarize both sides;
 cover the objection or someone else gets that citation.]

## [Comparison sub-question, if relevant: "X vs Y"]
[A fair table beats adjectives. State the cases where the competitor wins.]
```

Add FAQPage schema only when the visible content matches it. Generator: gtmoperator.org/schema-generator.

---

## 4. The content chunking template

AI retrieval works at the passage level (100–300 words), not the page level. Every section of a page should survive being read alone.

**Per-section checklist:**
- [ ] The heading is a question or claim, in buyer language
- [ ] The first sentence answers or states it directly
- [ ] Entity names appear in the section (the company, the product — not pronouns)
- [ ] Scope is explicit ("for B2B SaaS," "as of 2026") where it matters
- [ ] The section makes sense with everything above it deleted

**The pronoun test:** read any section alone. If "it," "they," or "the platform" is ambiguous, name the entity.

---

## 5. The llms.txt starter

Place at yourdomain.com/llms.txt — a plain-text guide for AI crawlers.

```
# [Company]

> [Canonical one-sentence description.]

[Two or three sentences of factual context: what the product does, who uses it,
 founding facts. No marketing language — this file is read by machines that
 score claims, not vibes.]

## Key pages
- [Homepage]: [URL]
- [What is / definition page]: [URL]
- [Pricing]: [URL]
- [About / founder]: [URL]
- [Research / data]: [URL]

## Facts
- Founded: [year]
- Founder(s): [names]
- Category: [category]
- Customers: [verifiable scale claim only]
```

---

## 6. The monthly question panel (for the tracker)

Fix 10–15 questions and never change them — the trend is the data. Mix:

- 3–4 definitional: "What is [category]?" "What does [company] do?"
- 4–5 recommendation: "Best [category] for [segment]?" "What should a [buyer] use for [job]?"
- 2–3 comparison: "[You] vs [competitor]?" "Alternatives to [incumbent]?"
- 2–3 problem-first: "How do I solve [the pain you fix]?"

Log results monthly in the AI-Visibility Tracker: gtmoperator.org/schema-generator (download section).

---

*The GEO Kit: audit (gtmoperator.org/geo-scorecard) → playbook (gtmoperator.org/geo-playbook) → these templates → the tracker. Questions: jgilbert@gtmoperator.org*