The creator-campaign measurement checklist

1. Define the decision the campaign should inform

Begin with the decision, not the dashboard. “Get exposure” is too vague to evaluate. A useful objective identifies the audience, action or learning:

• Awareness: reach relevant viewers efficiently with a disclosed, legible brand frame.
• Consideration: observe whether the placement and creator mix generate response without harming channel health.
• Traffic or acquisition: add an explicit path that can be attributed and define the attribution window.
• Creative learning: compare approved rail variants while holding other conditions as stable as practical.
• Market research: use a bounded pilot to estimate creator availability, clearing vCPM and delivery quality.

2. Name the billable unit and source

For a FrameYield Frame-Only placement, the billable unit is the validated YouTube thumbnail impression. The source is a connected YouTube Reporting API reach report, reconciled to the accepted video ID, date range and approved creative state.

The brief should name:

• the raw source field;
• the platform’s visibility definition;
• eligible and excluded surfaces;
• reporting lag and timezone;
• the validation window;
• invalid-activity and dispute handling.

If an agency proposal says only “impressions,” ask whether those are thumbnail displays, video ad impressions, social reach estimates or something else.

3. Define the inventory before seeing creator names

A recognisable channel is not automatically the best campaign fit. Set the rules first so selection is consistent:

• content category and prohibited adjacencies;
• target regions and language;
• standard or Established Creator cohort;
• minimum recent eligible reach and publishing cadence;
• audience-quality and brand-safety review;
• creator frequency cap and category exclusivity;
• number of approved backups if a creator declines.

Keep subscriber count as context. Use recent eligible thumbnail reach and content fit for the working forecast.

4. Freeze the creative evidence trail

The report should be able to answer “Which artwork delivered?” without relying on a screenshot in someone’s inbox.

• campaign rail percentage and side;
• brand variant name, dimensions, version and file hash;
• final 1280×720 FrameLock version and creator/brand approvals;
• required local disclosure wording;
• accepted YouTube video ID;
• published-thumbnail observation and comparison status;
• creative replacement or pause events.

Hashing alone does not solve platform recompression. The system should distinguish an exact source match from a high-similarity observation that still requires review.

5. Decide whether conversion attribution exists

A thumbnail does not naturally contain a standard clickable ad unit. If the campaign needs traffic or conversion evidence, choose and disclose a separate response path:

• creator-specific vanity URL shown in the sponsor rail;
• creator-specific promo code;
• description link or pinned comment, if separately accepted;
• brand-provided conversion event or aggregate report;
• controlled brand-lift or incrementality study.

For every signal, write down its owner, attribution window, deduplication rule, consent requirements and known gaps. If no signal is included, the conversion result is unknown. It is not “zero,” and it is not inferred from thumbnail CTR.

6. Set economics and stop conditions

A controlled pilot should have a budget, accepted vCPM range and maximum campaign value. It should also define what happens when:

• delivery reaches the cap early;
• the creator’s CTR guardrail is breached;
• the public thumbnail no longer matches the approved frame;
• reporting is delayed or the OAuth connection is revoked;
• the campaign ends with unused reserved budget;
• a brand or creator disputes an exclusion.

FrameYield’s beta uses a seven-day validation window and an 80/20 creator/platform split on validated gross media spend. Those terms belong in the accepted offer and campaign order—not only on a marketing page.

7. Reconcile the post-campaign receipt

Before reporting results, verify that every placement can connect these records:

1. Campaign order and reserved budget.
2. Creator offer and acceptance.
3. Brand rail variant and FrameLock approval.
4. Accepted video ID and active dates.
5. YouTube reach report rows.
6. Exclusion and validation reason codes.
7. Gross media value, creator payable, platform share and unused balance.
8. Optional attribution source and its separate limitations.

8. Write conclusions at the strength of the evidence

One-page copyable checklist

1. Define one primary business objective: Choose awareness, consideration, traffic, acquisition or research. Do not ask one impression metric to answer every objective.
2. Name the billable unit and source: Record the exact event, platform field, eligible surfaces, reporting lag, campaign window and exclusions.
3. Lock inventory and creative controls: Set creator cohort, category, geography, rail size, variants, rights, disclosure and approval gates.
4. Separate delivery from attribution: Decide whether the campaign is thumbnail-only or includes a creator URL, code, description link, event feed or lift study.
5. Reserve a bounded pilot budget: Agree the vCPM, maximum value, settlement window, unused-budget treatment and pause rules before launch.
6. Reconcile the post-campaign receipt: Match source data, video IDs, creative versions, exclusions, ledger postings and payment states before drawing conclusions.