Most campaign reports show ten metrics, none of which answer "did it work?" The cure is brutal selection: three metrics, no more.
Primary metric: the one tied to the goal. If the goal was downloads, the primary metric is downloads attributable to the campaign window. If the goal was RSVPs, it's RSVPs. Pick this before launch. Changing it after is how teams accidentally declare campaigns successful that weren't.
Leading indicator: engagement that predicts the primary. Saves and thoughtful comments in the first hour after each post correlate with distribution and downstream conversion better than likes. Track these per post so you know which hook is doing the work.
Lagging indicator: pipeline or revenue signal. A great campaign that earns 10,000 impressions but moves zero pipeline opportunities is not a great campaign; it's an awareness experiment. Tag campaign-sourced leads in your CRM so you can read this honestly 30 days later.
What to ignore: total likes, total impressions, follower growth during the window. They're not zero, they're context. But they're not the scoreboard.
The most useful artifact at the end of a campaign is not the report; it's the winner list. Which three posts drove the most primary-metric outcomes? Save them. Their hooks, their structure, their CTAs. Those become the seeds for your next campaign.
In SocialLens AI, this loop is automated: engagement signals from shipped posts feed back into the next campaign's draft. Hooks that worked once get prioritized; hooks that fell flat get deprioritized. The system gets sharper at writing for your specific audience over time, which is the real compounding move.
One last thing: report the loss honestly. Most B2B campaigns underperform the first plan. The teams that improve over the year are the ones that name what didn't work, not the ones with the prettiest deck.