Vanity metrics: numbers that feel good and mean nothing
Impressions are up, engagement is strong, CTR beat the benchmark, and none of it tells you whether marketing made you money.
Some numbers exist to be presented. They rise reliably, they chart beautifully, and they say almost nothing about whether your marketing worked.
The usual suspects. Impressions grow whenever you spend more: they measure your budget, not your marketing. Reach is impressions wearing a different hat. Engagement counts thumbs, not customers. Click-through rate feels like a quality signal, but a high CTR on the wrong audience is just efficient waste. Even "conversions" can be vanity when the conversion being counted is a scroll-depth event or a newsletter signup nobody reads.
Why they persist. Vanity metrics survive because everyone in the chain benefits from them. Platforms surface them because they always go up with spend. Agencies present them because they're defensible. Teams report them because they're available. Nobody is lying, but a reporting ritual built on numbers that can't go wrong is a ritual, not a measurement.
The test. For any metric, ask two questions. If this number doubled and nothing else changed, would the business be better off? And: could I make this number go up while making the business worse? Impressions fail the first. CTR fails the second: narrow your targeting to people who love clicking and never buy, and watch it soar. Metrics that pass both tests are rarer and harder: qualified leads, sales, revenue, cost per actual outcome.
What to look at instead. Anchor on the figures closest to money: what you spent, what you got that the business genuinely counts as a win, and what each win cost. Everything else is diagnostic: useful for explaining why the anchor numbers moved, useless as a verdict on whether they should.
A practical reframe: vanity metrics aren't evil, they're intermediate. Impressions explain a traffic change. CTR helps debug an ad. The failure is promoting them from explanation to scoreboard.
This is also a useful lens on any report you receive, from a platform, an agency, or a tool. If it leads with reach and engagement, ask what it's not leading with. OmniCommand takes a side here: every verdict it gives is anchored to spend and outcomes (the money in and the results the data actually recorded), with the intermediate numbers kept where they belong, as evidence. If a number can't be tied to something that happened, it doesn't get to headline your week.
Request early access at omnicommand.ai.