Marketing has a hoarding problem. Activities accumulate, the newsletter, the socials, the sponsorship, the directory listing, the event you always do, and nothing ever leaves, because every activity has a defender and none has a trial date. Budgets grow by addition and shrink only under duress. So here’s the test I’d apply to every recurring line in a marketing plan: if we stopped this, what would change, and when would we notice? If nobody can answer, stop it and find out.
The absence test beats the presence test
Most marketing measurement tries to prove presence: this activity produced these numbers. That’s fine where attribution is clean, but marketing attribution is rarely clean. The absence test comes at it from the other side. You don’t have to prove an activity works; you have to be able to say what its removal would break, and roughly when the break would show. Pipeline thins in a month. Enquiries dip by autumn. The brand weakens over a year, visible in these signals. Any of those is a real answer. “We’ve always done it” is not.
What switching things off can reveal
Businesses that run the experiment tend to find three categories. Some activities are load-bearing: stop them and something measurable sags, so they’ve earned their place and, usually, more investment than they were getting. Some are slow-burn: their absence shows up late but genuinely, which means they’re real but were being judged on the wrong clock. And some are zombies: activities whose removal changes nothing except the workload, kept alive by habit and the vague fear of stopping. In my experience the zombie share is greater than most businesses expect, and the time they consume can be worth more than the money they cost.
The brand objection, answered properly
The standard defence of unmeasurable activity is “it’s brand work, you can’t measure it.” Half right. Brand runs on a slow clock, and judging it in a quarter is exactly the horizon mistake boards make in the other direction. But slow is not the same as unmeasurable. Real brand work has believable long-term signals: awareness you can test, share of the conversations that matter, the rate at which people arrive already convinced. If an activity claims the brand defence but can’t name a signal or a horizon, it isn’t brand work. It’s a zombie wearing brand’s coat.
How to run the cull without wrecking anything
You don’t need courage so much as sequencing. List every recurring activity. Ask the absence question of each, and write the answers down, because writing exposes vagueness that conversation lets slide. Anything with no answer goes on trial: pause it for a defined window, watch the signals its defenders claim, and let the evidence decide. Protect the slow-clock work explicitly so the cull doesn’t eat the compounding assets. What survives will be a smaller plan, but done right, and the smaller plan can perform better, because the freed up time and budget flow to the activities that were quietly carrying everything.
Marketing effectiveness is usually framed as doing better things. Just as often it’s about ceasing to do meaningless ones. The absence test can find them. All it costs is the willingness to ask a question your busiest activities might fail.
