Here is the campaign lifecycle most brands still run: a brief, six weeks of production, a launch, and a report. The report is thorough. It is also useless, because by the time it lands the media budget has been spent and the team has moved on to the next brief.
We call the alternative closed-loop marketing. It is not a new channel or a new format. It is a different shape for the same work: strategy, production, distribution and measurement connected in a circle, so what you learn in distribution changes what you produce next week, not next quarter.
The Monday rule
The simplest test of whether a loop is closed: can what the data says on Monday change what is on air by Tuesday? If the answer is no, you have a line, not a loop. You have a campaign that will be judged once, at the end, when nobody can act on the verdict.
Getting to yes needs four things to be true at the same time, and most of them are organisational rather than technical.
- Creative made in variants, not as a hero. One film cannot learn. Twelve cuts with different hooks, different first three seconds and different offers can. Production has to be planned for iteration from the start, which means shooting more options than the storyboard needs.
- One set of numbers. Creative and performance usually look at different dashboards and argue about whose is right. If both teams see the same reads, the argument turns into a decision.
- A decision cadence with owners. Weekly is the right rhythm for most paid social. Someone has to be allowed to kill a cut, move budget and brief a replacement without a deck.
- Tolerance for losing your favourite. The cut the room loved will sometimes lose to the one nobody expected. The loop only works if the number wins.
What it looks like in practice
When we worked with Cakap on user acquisition, the brief was blunt: more registrants, lower cost per install. The loop was the strategy. Instead of one polished ad, we made native-language creative for specific communities, each with its own trigger, and read the results weekly. Cuts that did not move were retired without ceremony. Cuts that did were pushed harder and spawned new variants. Registrations doubled, and the weekly purchase uplift held because the learning never stopped.
A campaign that cannot change is a bet. A campaign that changes every week is a system.
Where loops break
Two agencies, two dashboards. Data that arrives monthly because a partner exports it by hand. Approval chains slower than the feedback. Creative teams who are told the results but never see them. Each of these turns the loop back into a line, and each is fixable, though rarely by adding another tool.
For our own clients we eventually built the plumbing ourselves. Cirya puts marketing, stock and sales in one dashboard on the client’s own servers, so the Monday read and the Tuesday decision happen in the same place. But the tool is the last step. The first is deciding that a campaign is allowed to learn.
A short checklist
- Is the creative planned in variants, with a budget for making more?
- Do creative and performance see the same numbers, at the same time?
- Who can move budget this week, and what evidence do they need?
- When did a campaign last change because of a number, rather than a meeting?
If the last answer is “never”, that is the place to start. Send us the brief and we will bring the loop.
