Every platform hands you a dashboard, and every dashboard quietly implies that bigger numbers mean it's working. But each metric answers a narrow question, and treating them as interchangeable leads to content strategies optimized for the wrong thing.
A field guide to the numbers
- Impressions answer: did the algorithm distribute this? They measure the platform's decision, not your audience's. A spike often says more about the feed's testing behavior than your writing.
- Likes answer: was this agreeable? A like is the cheapest possible action — acknowledgment, not endorsement. Safe, mild content collects likes efficiently. That's a warning as much as a win.
- Comments answer: did this provoke a response worth typing? Much stronger signal, with a catch: "Great post!" and a three-paragraph disagreement count the same in the total. Read the comments, don't count them.
- Followers answer: did someone bet on your future content? A follow is a subscription decision, which makes it more valuable than any single-post reaction — and the number everyone checks least often.
- DMs, replies, and mentions answer: did this move a relationship? Invisible on the dashboard, and where consulting clients, hires, and partnerships actually come from.
The inversion most people miss
For anyone using social media for business — rather than as the business — the value of the metrics runs roughly opposite to their visibility. Impressions are the most visible and least valuable; a DM from the right person is invisible and might be worth the whole quarter. If a post gets modest numbers but generates two real conversations, that post worked.
The dashboard measures what the platform can count. Your results live in what it can't.
What to do instead of staring at the graph
Pick a metric that matches your actual goal, and judge posts only against it. Building an audience? Watch follows per post, not likes. Generating pipeline? Track conversations started, in a note, by hand. Testing positioning? Read the specific words commenters use back at you — how people describe your idea is free market research on the idea itself.
And give any experiment enough runway. Single posts are noisy; the same post can perform differently depending on the hour it shipped. Judge patterns across ten posts, never verdicts from one.