Most activations are approved as events and then asked to behave like systems. Decide which one you are funding before the brief leaves the building.
The market commentary behind this piece puts the figure at 73% of brand activations failing to drive meaningful business outcomes. Treat that as a directional claim rather than a measured industry rate. The useful part is the stated reason: activations get designed as isolated events and then judged as if they were infrastructure.
A campaign and an activation are different commitments
A campaign has an end date, a fixed asset list and a reach target. An activation is a mechanism you intend to run again, in another market, with different partners and the same core structure. The two need different budgets, different approval gates and different definitions of success.
Write down which one you are buying before creative work begins. If the answer is an activation, the brief must name the parts that travel, the parts that get localised, and the person accountable for each in every market.
If nothing is intended to run twice, call it a campaign and stop paying for the systems work it does not need.
What makes an activation repeatable
Repeatability is a property of the components, not of the idea. Four conditions are worth agreeing before the first market goes live.
Fix the parts that never change: message, structure, quality floor.
Define exactly what a market may adapt, and what it may not.
Name the owned, earned and paid role of every touchpoint.
Decide which signals return to the kit, and who may edit it.
Agree all four before the second market, not after it.
Four components that decide whether an activation can be repeated: core kit, local layer, channel orchestration and a feedback loop.The physical and digital sides have to be specified together. If the QR code, the messaging thread and the in-person moment are scoped by different teams, the join between them is where the experience fails and where measurement disappears.
Modularity has a cost. Building components that survive translation is slower than producing one bespoke event, and that cost only returns if a second and third market are actually committed.
Where the delivery model breaks
Four patterns recur in agency-delivered activations, all of them structural rather than creative.
- 01 Campaign thinking applied to system work, so nothing is built to be reused.
- 02 Reach treated as the outcome, which optimises for impressions over participation.
- 03 Platform dependency, which hands the performance of your activation to somebody else’s algorithm.
- 04 Creative settled before strategy, producing work that is memorable and unattributable.
None of these are arguments against agencies. They are arguments for specifying the operating model in the contract: who owns the component library, who may change it, and what the acceptance criteria are per market.
Measure outcomes, not reach
Reach tells you the activation happened. It does not tell you whether anything changed. Agree the measurement set before launch, because retrofitting it means reconstructing decisions nobody logged.
Cheapest to report, weakest evidence. Use for pacing only.
Did people take the action the activation was designed around?
Did the same people come back without being paid to.
Revenue, retention or cost avoided, attributed to the mechanism.
Each step costs more to instrument and carries more weight in a budget review.
Four levels of activation evidence, from impressions through participation and return to commercial outcomes.Set a threshold on the second and third measures before the first market runs. A number agreed in advance is what separates a scaled programme from a repeated expense.
Decide what to test next
Start with the smallest thing you can approve: one market, one channel, one defined participation metric, four weeks. The source case for this piece describes ten cities in four weeks using messaging, QR codes and local creator networks. Read it as an illustration of sequencing, not as a benchmark to hit.
Before expanding, state what would make you stop: participation below the agreed floor, a market that cannot run the kit without rebuilding it, or a channel whose economics only worked while it was subsidised.
An activation should be assessed against a defined mechanism and a named owner, not inferred from a successful launch event.
Written by the studio, from work in delivery
