One question: if this number moved, would I do something differently?
If the honest answer is no, it is not a metric, it is decoration. Follower count usually fails this test. So does impression volume, time on page, and number of posts published.
Why weak metrics survive
Not because anyone is fooled. Because they go up reliably, they are available without effort, and reporting them is safer than reporting a number that might be flat. The incentive is organisational, not analytical.
Easy to grow, weakly connected to revenue, and comparable across companies, which is exactly why it gets reported. A follower who will never buy costs you nothing and earns you nothing.
Look at instead: qualified enquiries that came from that platform.
Counts how many times something was displayed. It is possible to double impressions and halve revenue, and this happens routinely when reach is bought cheaply from the wrong audience.
Look at instead: reach among your actual target, or just skip to enquiries.
Frequently misread as failure. If someone lands on a page, gets their exact answer, and leaves satisfied, that registers as a bounce. For reference content, a high bounce rate can mean the page worked.
Look at instead: for a page meant to convert, the conversion rate. For a page meant to inform, whether it produces branded searches or return visits later.
Too aggregated to act on. It mixes people who searched your name with people who saw a cold ad. Those two groups want completely different things. The average tells you nothing about either one.
Look at instead: conversion rate by source, and by step.
The comparison that misleads most
Comparing your conversion rate to an industry benchmark. Benchmarks rarely tell you their traffic mix. Someone who searched your brand name is nothing like someone who saw an ad. Same metric, different people. Compare yourself to your own previous numbers.
Five metrics each. Resist adding more; the value of a dashboard falls as it grows.
Qualified enquiries per month
Percentage of enquiries that become proposals
Percentage of proposals that close
Average project value
Self-reported source of enquiry
The two conversion steps are separated deliberately. Losing people at the enquiry stage and losing them at the proposal stage are completely different problems.
The single number closest to the thing you are currently trying to fix. Nothing else.
Monthly: the five-metric dashboard
Enough time for patterns to be real rather than noise.
Quarterly: cost per customer and payback
These move slowly and are distorted by short windows.
Never: everything, every day
Daily monitoring of a full dashboard produces reactions to randomness, which is worse than not looking.
On small numbers
Most small businesses do not have the volume for small differences to be meaningful. Two conversions versus one is not a doubling, it is noise. Before acting on a difference, ask whether a coin flip could have produced it.
