Every platform dashboard is confident, colorful, and subtly incomparable. Each counts differently, defines differently, and rounds in its own favor. The result, for a brand publishing in more than one place, is a wall of numbers that feel like a report and behave like a rumor.
The fix is not a bigger dashboard. It is knowing which numbers survive being compared across platforms, which ones only make sense in context, and where the honest trail ends.
The metrics that travel
A handful of measures mean roughly the same thing everywhere, and those are the ones worth a cross-platform view: impressions or views (with each platform's definition read once, then held constant), engagements as a family — likes, comments, shares, saves — clicks to your own property, and watch time for video.
Even these deserve suspicion at the edges: a "view" on one platform starts at three seconds and on another at a single impression. The practice that makes comparison honest is preserving the raw values each platform reports and normalizing only at the last moment, visibly — not baking a blended "engagement score" that quietly averages unlike things.
The metrics that lie
Follower count lies by direction: it is a stock, not a flow, and it moves for reasons (a viral moment, a purge, a platform promotion) that say nothing about the content you shipped that week. Reach lies by selection: platforms show it for their own preferred formats, so a "reach win" can mean the format was favored, not the message. Composite scores lie by construction: any single number that blends engagement, sentiment and growth has made decisions you did not sign off on.
None of these are useless — they are context. They become lies when stacked into a single ranked leaderboard across platforms.
- Follower count — a stock, not a signal of this week's content.
- Platform-reported reach — favors the platform's preferred formats.
- Composite "engagement scores" — blends unlike definitions into one confident number.
The question behind the metrics
The metrics exist to answer one question: did this content move the business? The full trail runs content → click → landing → lead → revenue, and every arrow after the first lives partly outside the platforms' walls. Platforms report what happened on the platform; your analytics report what happened after the click; joining them is your job (or your tool's), and it is never perfect.
Being honest about the seam matters more than papering over it. Attribution windows differ, last-click flatters the closer and robs the opener, and some conversions have no clickable ancestor at all. Report the trail with its gaps, and the numbers keep their meaning.
A reading habit that works
Weekly, read three things and resist the rest: engagement rate per post within its own platform (against that platform's own baseline), click-through to your property, and watch time or read-through for the long pieces. Monthly, add the trail: which pieces produced clicks, which clicks produced leads, which leads produced revenue — with the seams visible.
Everything else — the follower odometer, the reach leaderboard, the blended score — goes in the drawer labeled "context", opened when you have a specific question it can actually answer.