✓ You've read this lesson
3.2The report we want to show, and resonance
Write the report’s headlines before you pick a single creator. Decide what you’ll show, what you want to learn, and what you’ll compare against.
Metrics by behaviour
| Core Impact (always) | Upgrade (needs) | Cost efficiency | |
|---|---|---|---|
| Stream | Viewers who reached the brand segment, and how much of it they watched | Brand uplift study (+ media partner) | Cost per segment view |
| Scroll | Resonance, plus unique reach | Brand uplift study (+ media partner) | Cost per engagement, cost per share |
| Search | Branded search lift, plus target query presence | Organic search traffic to the brand’s site (+ site analytics); AI answer presence (+ tooling) | TBC: too early to set a cost target |
| Shop | Attributable sales, creator traffic conversion | Retailer sell-through (+ retailer data); direct and referral traffic lift, for brands selling online (+ site analytics) | CPA, ROAS trend |
Core metrics must go in every report. Add upgrades when the budget or tooling allows. Only ever compare cost efficiency with other creator activity. Definitions are in the appendix.
Goals, and what we want to learn
Set the goals, and one thing to learn, before launch. Goals tell you whether the campaign worked. What you learn makes the next one better.
A goal is what you want to achieve: resonance above 1.2, or cost per engagement 10% below the last campaign. What you want to learn is a question with a comparison built in, because most campaigns are too small to prove anything on their own.
Checklist
- One Impact and one Cost Efficiency metric
- A baseline named for each
- Goals written as numbers
- One thing to learn, with its comparison
Resonance: the metric on every campaign
Resonance asks one question. Did the audience lean in to the branded post more, or less, than they do to this creator’s normal content?
Weight each interaction by the effort it takes: a share or save counts more than a comment, and a comment more than a like. Score comments for relevance to the advertiser’s product, and count only the relevant ones. Report the sentiment of those comments alongside the score, because a relevant negative comment is a signal to act on, not something to hide. Then divide the weighted score of the branded post by the same weighted score for the creator’s median post.
Below 1.0
1.0
Above 1.0
What weighs most (illustrative order, not final weights)
- Shares and sends
- Saves
- Relevant comments
- Likes