How we review creator media kit tools, and where the documentation ends

Creator media kits review methodology should define the evidence needed for a verdict before a reviewer opens the product. A source page can support an account of what a provider documents. It cannot establish the experience of making, maintaining or delivering a creator portfolio. Use the proposed protocol below to keep those findings separate.

What to take away

Specify the review commission

For this protocol, the commissioning publication is creatorportfolio.co.uk. No independent reviewing body has been appointed, and the named human reviewer and credentials remain unassigned. This is AI-assisted methodology drafting based on sources examined on 6 September 2026, not an executed product trial.

The subject of a future review must be a named media-kit product and identifiable configuration. Record the plan, application version where available, device, browser, review date and exact kit task. If a hosted service exposes no version number, record that limitation instead of inventing one.

Write the intended decision, such as whether a candidate can support a specified portfolio handover. Keep unrelated outreach, payments and campaign-management functions outside the commission unless separately authorised and researched.

Assign a source class to each statement

Mark provider help pages and sales descriptions as first-party evidence. Mark independently collected observations separately, including who collected them and under what conditions. Do not treat a testimonial displayed by the provider as independent verification of the proposed workflow.

For example, Canva's flattened-PDF instructions document an export route. Reading that page establishes what the help record describes; it does not show what happened to a particular kit's links, text or accessibility. A later reviewer would need to inspect the actual output to address those questions.

Maintain a claim log with the source URL, access date, affected statement and evidence boundary. A changed or unavailable page needs rechecking before release, even if the earlier draft cited it correctly.

Worked claim-log entry:

Design the task before running it

Use material the reviewer is authorised to handle and define the requested receiving format. Specify which content changes must be attempted, which output must be inspected and how another person could reproduce the steps. Decide in advance what constitutes an unresolved result.

For trial material, GOV.UK's copyright page sets out ways to obtain permission and circumstances in which permission is not needed. It does not clear a specific image for a trial. Obtain the necessary rights decision for actual portfolio samples rather than assuming research use permits copying everything a creator has published.

Separate a planned activity from a completed observation in the log. Record the actual time and result only after performing the step. Do not populate blank fields with estimates of how an experienced user might fare.

Evaluate the receiving copy

Ask whether the proposed reviewer can locate the work samples, understand their captions and follow the intended navigation in the exact output. Retain any relevant settings and limitations. A successful action in an editing workspace would not by itself settle the receiving-copy question.

Government accessible-document guidance supplies principles for document structure and checking. Its government-publishing context must remain visible when applying it as a house review safeguard. It neither certifies a private kit nor establishes that a product has passed accessibility testing.

Where specialist assessment is needed, name the reviewer and method before making a conformity claim. Leave unperformed checks untested, without converting them into either a pass or a product failure.

Write a verdict proportional to the record

Summarise the tested task and configuration before stating a finding. Distinguish a documented feature, an observed behaviour and a judgement about suitability. Explain what remains unknown and whether it prevents the intended selection decision.

This draft involved no account use, subscription, file creation, export, support contact or user feedback. It has no supplied trial or affiliate recommendation; publisher relationships still require confirmation. Consequently, it supplies a review procedure and withholds all product-performance conclusions. Complete named authorship, fact-checking and methodological review before publishing the protocol or commissioning a claim-bearing review from it.

Before you act

Common questions

What should a reviewer do if a cited source page changes or becomes unavailable?

A changed or unavailable page needs rechecking before release, even if the earlier draft cited it correctly. Maintain a claim log with the source URL, access date, affected statement and evidence boundary to track such changes.