AI Search Content Brief Template: SEO, AEO and GEO

By Abiot Y. Derbie - Founder, RankEcho - Updated 2026-09-07

# AI search content brief

Manual template — version 1.0. Complete the required fields and delete the instructions before handoff. The brief makes decisions and evidence reviewable; it does not predict a ranking, AI appearance, citation, visit, or sale.

## 1. Brief control
- Brief title [required]:
- Status: Draft / In review / Approved / Published / Superseded
- Canonical page URL [required]:
- New page or refresh [required]:
- Page type: Guide / Product / Pricing / Comparison / Use case / Support / Other
- Brief owner [real name required]:
- Writer [real name or TBD]:
- Subject-matter reviewer [real name or TBD]:
- Editorial reviewer [real name or TBD]:
- Product, legal, security, or compliance reviewer [name, not required, or TBD]:
- Target publication date:
- Brief version and last update:

## 2. Reader, job, and intent
- Primary audience [required]:
- Situation or trigger:
- Decision or task the page must help complete [required]:
- Primary query [required]:
- Natural conversational question [required]:
- Related questions:
- Market and language [required]:
- Buyer stage:

## 3. Search and source context
- Review date and conditions:
- Queries sampled:
- Representative result URLs:
- Representative sources visible in sampled AI answers, if checked:
- What readers need immediately [required]:
- What existing results answer well:
- What is missing, unclear, unsupported, or hard to use:
- Sample limits: qualitative review only; not search volume, difficulty, market share, or a selection explanation

## 4. Ownership and differentiation
- Existing canonical owner:
- Neighboring owner → distinct job:
- Why this page deserves a separate URL [required]:
- Questions handled elsewhere → owner URL:
- Pages to consolidate instead of duplicating:

## 5. Reader outcome
- Direct answer the page owes [required]:
- Publicly useful artifact or result [required]:
- Information advantage: original example / tested method / first-party evidence / useful synthesis / other
- Explicit non-goals and unsupported claims to exclude:

## 6. Buyer-question map
| Buyer question | Section | Answer required | Evidence required | Source | Reviewer | Status |
| --- | --- | --- | --- | --- | --- | --- |
| [Question] | [H2] | [Specific answer] | [Fact, method, example, or none] | [Exact URL or pointer] | [Name or TBD] | Open |

## 7. Claim, source, and freshness register
| Exact proposed claim | Type | Source | What it supports and does not support | Checked | Freshness owner and trigger | Named reviewer | Status |
| --- | --- | --- | --- | --- | --- | --- | --- |
| [Claim] | First-party fact / external fact / inference / example | [Exact source] | [Scope and limit] | [YYYY-MM-DD] | [Owner + event/date] | [Name or TBD] | Verified / Needs source / Revise / Remove / Synthetic-only |

Rules:
- A URL alone is not a review; record its support and limits.
- Keep as-of dates beside changing prices, features, policies, and comparisons.
- Do not publish Needs source, Remove, or Synthetic-only material as real fact.
- A role placeholder is not completed approval. Record a real reviewer or leave TBD.

## 8. Page plan
- Opening answer:
- Title:
- H1:
- Meta description:
- Canonical:
- Section → question → answer → claim/source IDs → artifact:
- Internal links: parent hub / previous task / next task / evidence / honest product destination
- Applicable structured data: only if it matches visible content; no special AI schema claim
- Conversion handoff: label / destination / eligibility / what it does / what it does not do

## 9. Delivery and review
- Primary answer and artifact in accessible public HTML: Yes / No
- Canonical, status, robots, sitemap, and index policy checked: Yes / No
- Title, H1, description, visible copy, and applicable schema agree: Yes / No
- Mobile and keyboard use checked: Yes / No
- Material claims verified by actual named owners: Yes / No
- Unsupported material removed or visibly unresolved: Yes / No
- Final URL, body version, reviewers, and UTC ship time recorded: Yes / No

## 10. Measurement and decision
- Baseline, window, units, denominators, exclusions, and concurrent changes:
- Traditional search discovery:
- Provider-native AI reporting, with each provider's unit kept separate:
- Fixed prompt-and-surface observations, including unavailable and failed checks:
- Identifiable AI referrals and landing pages:
- On-site task use and key events:
- Commercial outcomes only where a reliable join and attribution rule exist:
- Decision: Keep / Improve / Consolidate / Retire / Investigate product friction
- Interpretation: observed after publication / attributable / assisted / unknown

## Unresolved gaps and publication decision
| Gap | Owner | Resolution needed | Status |
| --- | --- | --- | --- |
| [Gap or none after review] | [Name/role] | [Source, decision, test, or approval] | Open / Resolved |

Publication decision: Ready / Not ready
Decision owner [actual person]:
Decision date:
Reason:

How should you use this content brief?

Copy the blank brief before drafting. Complete the reader, intent, ownership, question map, evidence, freshness, structure, delivery, conversion, and measurement fields. If a required fact has no suitable source or reviewer, leave it visibly unresolved and keep it out of publishable copy.

Use one brief for one canonical page task. Closely related query and conversational phrasings can share a page when they serve the same decision. A new wording variation is not, by itself, a reason to create another URL.

This is a manual editorial tool. It does not query a search provider, estimate demand, import RankEcho data, approve a claim, or publish anything.

Which fields belong in an AI search content brief?

SEO, AEO, and GEO do not require three duplicate briefs. The useful distinction is between the reader's task, the answer the page owes, the evidence needed to support it, and the work required to ship and review it.

How do buyer questions become page sections?

Map each distinct question to the section that owes the answer. State the answer required and the evidence needed before assigning prose. Merge questions that support the same decision; route materially different jobs to their established owner.

A question heading, concise answer, table, or list may improve usability when the information suits that form. None is a universal selection rule, so choose the clearest representation for the reader.

What does a completed SaaS brief look like?

This filled example uses Harborline Cloud, an imaginary B2B SaaS company, and a proposed customer-onboarding analytics guide. The company, page, facts, records, roles, and approvals are synthetic. They are not customer data, market evidence, or publishable product copy.

The record is structurally complete but ends Not ready. It records that no live result review was run, blocks an unsupported setup-time claim, removes an outcome claim, and leaves actual reviewers to be named. A completed field set is not a fabricated approval.

How are the synthetic example questions and claims filled?

The excerpt below fills the production fields rather than treating a company name as a completed example. Harborline Cloud and every path, product fact, source pointer, role, and decision remain fictional. Synthetic-only and unresolved claims cannot move into publishable copy.

How should sources, freshness, and review be recorded?

Copy the exact proposed claim into the register. Link the specific supporting page or evidence artifact, explain what it supports and what it does not, add the date checked, identify who owns freshness, and state the event that should trigger another review.

Use Verified only after a suitable source and an actual reviewer support the wording. Needs source stays out of the draft, Revise returns for narrower wording, Remove is excluded, and Synthetic-only remains visibly fictional. A role placeholder is not completed review.

• Product and pricing facts: use the canonical owned source, an as-of date, and a product or finance reviewer.
• External method or market facts: prefer an appropriate primary source and preserve its scope and limits.
• Performance or customer outcomes: require the eligible population, denominator, dates, method, permission, and limitations.
• Inference: identify the supporting observations and label the interpretation rather than presenting it as measured fact.

What should you check before and after publishing?

Before publication, confirm one intent owner, public usefulness, complete claim records, actual reviewers, consistent title, H1, description and canonical, accessible primary HTML, working internal links, mobile and keyboard use, accurate structured data where applicable, and a CTA that describes its destination honestly.

After publication, keep traditional search discovery, provider-native reporting, fixed answer observations, identifiable referrals, on-site actions, and commercial outcomes separate. State each unit, denominator, exclusion, attribution rule, observation window, and concurrent change.

Describe movement as observed after publication unless the measurement design supports a stronger statement. A later ranking, citation, visit, signup, or sale does not by itself show that the page caused it.

How can RankEcho help after the brief?

The public sample fix shows one deterministic, fictional package shape before a paywall. RankEcho's paid or trial Fix Engine can turn an observed gap into a proposed bounded artifact, such as an answer passage, applicable schema guidance, page structure, or source-research plan. A person still decides whether its claims, sources, placement, markup, and deployment are accurate.

The Fix Engine does not automatically publish the page and does not promise discovery, ranking, citation, traffic, or conversion. Full-page writing and refresh work is a separate content add-on. This template remains usable without an account.

Frequently asked questions

Sources reviewed

Material technical claims below were checked against primary provider documentation. The sources support the documented control or signal, not a guarantee of indexing, ranking, an AI impression, or a citation.