Tools and providers
Part of How to assess software for influencer-agency selection
Influencer agency tool implementation: Traackr access and release checks
Assess an influencer agency's tool workflow with Traackr API access checks, ICO data protection guidance, CMA ad labels and explicit release tests.
Assess an influencer agency's implementation proposal against a defined task, not the polish of its demonstration. Ask the team to use fictional records, show its proposed permissions and preserve evidence of decisions. Agree what would justify release before giving the proposed configuration access to live campaign work.
Define the task and agency responsibilities
Use an agency's directory entry to prepare questions, then request evidence for the implementation task you actually need. Specify who prepares the record, who reviews it, who approves a change and who owns the final decision. Keep proposed responsibilities distinct from actions demonstrated during the test.
Choose a narrow task, such as moving fictional content through review, correction, approval and archive. Write down the required fields and expected decisions before the session. Keep software selection separate from implementation assessment: ask whether the proposed configuration completes your task, rather than accepting a general product presentation.
Check Traackr credentials before testing
The Traackr API PHP client needs an API key. Contact api@traackr.com to obtain one. Ask the agency to identify the proposed test actions before requesting access. Keep credential ownership, the authorised operator and the purpose of each request in the implementation record, alongside the agreed restrictions.
The client's bundled public API key can run tests only against read-only endpoints that do not add, modify or delete data. Do not treat a successful read-only test as evidence that a write action works. Record that boundary before assessing whether the agency has completed the agreed task.
Running the entire test suite requires an API key with full API access, which can be requested by emailing api-support@traackr.com. Ask the agency to distinguish completed tests from tests awaiting that access. Do not accept an untested write action as evidence of a working implementation.
Keep personal data out of evaluation
The ICO data protection guidance says organisations must collect only what they actually need and should not ask for or keep information 'just in case'. Use fictional creator records for evaluation, and ask the agency to justify each field proposed for the live workflow.
Organisations must tell people why they need their data, who they will share it with and how long they will keep it. Ask the agency to identify who owns those notices and decisions. Record those responsibilities alongside the data protection questions assigned to the buyer.
Keep the Data Protection Act 2018 and UK GDPR in the legal review brief, with unresolved questions assigned to an appropriate adviser. The UK GDPR includes a right to have inaccurate personal data rectified, or completed if incomplete. Ask how a correction would be reflected in the proposed workflow.
Test disclosure review before release
The CMA creator guidance requires incentivised content for an online audience to be clearly labelled as an advert or ad and based on genuine experience. Build that check into the proposed approval task, and ask who records the decision before the fictional publication step can proceed.
Labels must be clear, prominent and easy to understand. If a labelling tool does not clearly convey that content is an ad, include #Ad in a clear, upfront position. For carousels or sequences with individually accessible posts, label each item containing promotional messaging. Test those checks with fictional content.
The Advertising Standards Authority (ASA) enforces industry advertising rules. Include the Committee of Advertising Practice (CAP) and CMA guide on how to label ads correctly in the review. Its scope covers consumer protection law and the Advertising Codes enforced by the ASA; use advertising rules to organise questions.
Apply explicit acceptance conditions
Agree the following proposed decision rules before the test: PASS only when every required check has recorded evidence; HOLD when evidence is missing or an action remains untested; FAIL when the observed configuration contradicts an agreed restriction. Apply them to the specified task, not to the product as a whole.
| Check | Required evidence for PASS |
|---|---|
| Task completion | The fictional record completes every agreed state, with required fields and decisions matching a separate manual log. |
| Credential boundary | Each agreed API action is demonstrated using authorised credentials, within the permissions approved for the test. |
| Action traceability | Each tested system action, including an API call, can be associated with the person or account that performed it. |
| Disclosure approval | The fictional publication step cannot proceed until the required disclosure review and approval are recorded. |
| Access closure | Temporary test accounts are removed or suspended when no longer required, with the closure action recorded. |
Use the operating workflow to organise the states and hand-offs you want to examine. Record the observed result beside each condition, including the account used and the evidence retained. Keep a missing demonstration on HOLD; do not replace it with an assurance that the configuration should work.
Close access and agree handover
Remove or suspend temporary test accounts when they are no longer required. Include joiners, movers and leavers processes so access can be revoked when no longer needed or changed when roles change. Assign ownership of account closure before the exercise, and retain evidence that the agreed action occurred.
Ask the agency to describe its implementation responsibilities, the access it needs and the records it would hand over at the end of the relationship. Use vendor due diligence to organise wider supplier questions. Approve only the demonstrated scope, with unresolved defects, untested actions and further review responsibilities recorded.
