Understanding influencer campaign analytics

A dashboard helps you follow a campaign and identify questions to investigate. Its value depends on what metrics mean, their period and data freshness. A quickly updated counter does not guarantee complete attribution or campaign profitability.

Start with three questions

Which event is counted? How recent is the data? Which decision are you trying to make? Before comparing creators, check that campaign, dates, payment model and conversion type are comparable.

Tracking events, aggregates and financial operations can progress at different times. Separate an event reported by the website, a result accepted for compensation and the creator’s payout. Use the tracking guide when those stages disagree.

Read metrics without changing their meaning

• Clicks: clicks recorded on campaign links under the counting rules; not automatically unique people or qualified prospects.
• Leads and purchases: actions defined and reported by the integration. A sign-up and a sale are not interchangeable.
• Cost per action: included spend divided by the corresponding action count. Without an action, cost per action is not interpretable.
• Conversion rate: conversions divided by clicks in the same scope, with a consistent definition of conversion.
• Impressions and CTR: check whether impressions are measured or estimated. CTR calculated from an estimate does not certify a social network’s views.

Estimated revenue and ROI need reconciliation with store orders and actual costs. Our ROI guide separates revenue, margin and profit.

Follow budget and commitments

Compare committed compensation, recorded spend and remaining campaign budget. The wallet also separates available funds, reserved amounts and pending requests.

Performance pay may be supplemented by a creation fee. Consider all those commitments before making a change. Available actions depend on the collaboration’s state; a dashboard does not mean a creator can be freely replaced or an accepted commitment changed.

Investigate an anomaly methodically

If clicks rise without purchases, first test the link and landing page. Then check consent, the final event, order references, offer availability and the usual purchase delay.

A blocking rate or sudden variation is a signal to investigate, not sufficient proof against a creator. Keep context before contacting support: period, campaign, observed change and expected result. The fraud prevention guide explains that process.

Turn monitoring into decisions

Choose a review schedule appropriate to campaign duration and volume. Record observed figures, their limits and one hypothesis to test. Change a limited number of variables so that the result remains interpretable.

For a team review, prepare a dated record from available data and store orders. This method does not imply that every WeeGlad screen provides a particular export, custom alert or industry benchmark.