What this template does It turns methodology claims into explicit fields a buyer, agency, in-house team, or measurement provider can compare and preserve. A blank becomes [Not disclosed] in the output instead of disappearing from the record.
Independent implementation aid—not an official IAB tool The field structure was informed by the August 2026 IAB Measuring Visibility in the AI Era framework and Quoted First’s public repeatability work. This page is not produced, certified, endorsed, or reviewed by IAB. It does not determine compliance or whether data is fit for a business decision.
Browser-local generator

Build the methodology record

Write enough detail for another reader to reconstruct what was observed and identify what remains unknown. The 26 core fields are an omission inventory, not a pass/fail test.

01 · Identity and intended use

Name the method and the decision it supports

Use the owner and the named method, not a campaign slogan.
A date alone is not a version if the method can change.
The date this disclosure became accurate.
The generator does not verify this declaration.
02 · What is measured

Declare platforms, prompts, queries, and source classes

If no source classes or competitive metric are used, say so explicitly.
03 · Collection environment

Make each observation reconstructable

04 · Classification and calculations

Define every judgment and formula

05 · Validation, uncertainty, and maintenance

Show what makes comparisons credible—or only directional

Generate · inventory first, preserve next

The output keeps every field in a fixed order and labels blanks as not disclosed. Generation does not verify any answer.

Complete any fields you can. Empty core fields will remain visible as omissions in the generated record.

What belongs in a minimum usable disclosure?

A reader should be able to answer six questions without booking a sales call or guessing from a dashboard:

  1. What was measured? Named entities, platforms, models or surfaces, markets, languages, prompt set, date window, and exclusions.
  2. How was it collected? Architecture, access method, retrieval state, account/session conditions, cadence, repeats, failures, and denominators.
  3. How was it classified? Mention, citation, prominence, portrayal, recommendation, accuracy, and source matching rules.
  4. How was it calculated? Numerators, denominators, weights, normalization, bands, and every provider judgment.
  5. How stable is it? Repeated-run variability, uncertainty, sample size, platform/model changes, and re-baseline policy.
  6. Can it be checked? Raw answers, prompts, URLs, timestamps, nulls, versioned exports, retention, ownership, corrections, and known limits.

A complete-looking record can still be wrong

This tool counts whether a core field contains text. It does not verify that the text is accurate, specific, internally consistent, contractually binding, independently validated, or sufficient for a high-stakes decision. Treat a filled record as the start of review, not the conclusion.

Keep measurement layers separate A linked source, an unlinked mention, a recommendation, factual accuracy, sentiment, traffic, pipeline, and revenue are different observations. A disclosure should never fold them into one unexplained score or imply that movement in one proves movement in another.

Already have two reports with different numbers? Use the AI visibility score reconciliation template to compare their constructs, prompts, platforms, formulas, windows, uncertainty, and raw evidence without averaging unlike metrics.

Frequently asked questions

What should an AI visibility measurement methodology disclose?

Disclose what is measured, platform and prompt coverage, collection conditions, classification rules, formulas and denominators, invalid-run handling, validation, variability, uncertainty, raw evidence, historical versioning, retention, limitations, and corrections.

Does this generator certify an AI visibility provider?

No. It inventories disclosed information and produces a portable record. It does not test truth, completeness, decision-grade fitness, IAB compliance, certification, endorsement, legal sufficiency, or provider quality.

Why label blank fields instead of removing them?

An omitted field can materially change interpretation. Keeping [Not disclosed] in the record lets a buyer distinguish “not applicable,” “unknown,” and “not supplied” instead of assuming the most favorable answer.

Does the generator upload form data?

No. Generation, copying, and Markdown download happen in the browser. The page does not upload, transmit, or store the entered disclosure text.