Stop Measuring the Pulse, Start Treating the Patient: A Rant About Vanity Metrics

Every month, thousands of UK marketing managers receive a beautifully designed PDF from their agency. It's got charts. It's got upward-trending arrows. It's got a headline number in a nice big font: 'Impressions up 34% this quarter!'

And every month, somewhere in that same business, someone asks the awkward question: 'Lovely — but did we sell anything?'

Cue the shuffling of feet.

The Sugar Pill Problem

Vanity metrics work exactly like a placebo. They trigger a genuine feeling of improvement — the dopamine hit of a viral post, the warm glow of a follower milestone — without actually treating anything. The patient feels better. The patient is not better.

Here's the uncomfortable truth: a brand can gain 10,000 followers in a month and be in worse health than when it started. If those followers came from a giveaway or a trending audio with no connection to what you actually sell, you haven't grown an audience — you've grown an audience-shaped problem. Hollow reach inflates your costs (more followers means every future post needs to work harder) while diluting your signal to the people who might genuinely buy.

Impressions are worse still. An impression is not a person. It's not even a glance. It's a metric that counts the same doom-scrolling user three times because your video looped. Celebrating impressions is like a GP celebrating the number of times you walked past the surgery.

The Agency Incentive Problem

Let's name the elephant: plenty of agencies report feel-good numbers because feel-good numbers are easy to produce and impossible to argue with. Revenue attribution is hard. Retention cohorts are hard. 'Reach' is a number you can screenshot.

When an agency's contract renews on the strength of a quarterly deck, the path of least resistance is to optimise for whatever makes the deck look good. That's not malice — it's incentive design. But it means the metrics that matter most are the ones least likely to appear in the summary slide.

Ask your current agency one question: 'Show me the last pound of revenue social media directly generated, and how you know.' If the answer takes longer than a minute or involves the word 'awareness' used as a shield, you have your diagnosis.

The Metrics That Actually Treat

A proper metrics regime — one we'd happily put our name to — rests on three pillars:

Revenue attribution. Not the mystical 'last click' kind, but honest tracking: UTM-tagged links, promo codes unique to each channel, and platform-native shops where the path from post to purchase is measured end to end. You should be able to say, with evidence, what a follower is worth.

Retention and repeat behaviour. A customer who buys once from a viral post is a fluke. A customer who buys twice is a strategy. Track returning customers from social channels separately from one-off deal hunters, because the gap between them tells you whether your content builds relationships or just traffic.

Real conversations. The humblest metric of all: how many genuine, two-way exchanges did your brand have this week? Not replies — conversations. This is the leading indicator of everything else. Communities that talk to you buy from you, forgive you when you slip, and recommend you unprompted.

Breaking the Placebo Habit

If you're a brand owner reading this with a sinking feeling, don't bin your agency tomorrow. Do this instead: demand one page, every month, containing only commercial outcomes — revenue attributed, cost per acquisition, repeat purchase rate and conversation count. If those four numbers aren't trending the right way over two quarters, no amount of impression growth should save the contract.

And if you're an agency reading this: yes, this is a challenge. The UK market is maturing, clients are savvier, and the sugar pill is losing its effect. The firms that thrive in the next five years will be the ones comfortable being judged on the till, not the timeline.

The placebo makes the patient feel better. The treatment makes the patient be better. Choose your metrics accordingly — because your business can't build a recovery on numbers that were never real in the first place.