One Core Message per Scenario
In journey design, introducing a new core message at every touchpoint often leaves people unable to recall what they were meant to believe by the end.
Article AI Marketing Property & Tourism
Each experiment leaves behind one reusable judgment
When closing an experiment, if the records still only note whether results were good or bad, the next experiment usually can't borrow this one's judgment. A sounder approach: every close-out must state one judgment that doesn't depend on this round's materials, fixed in place with a judgment statement.
Key points
- At close-out, experiment records should no longer stop at whether the results were good or bad.
- Leave it alone and the same thing keeps happening: the next experiment can't borrow this one's judgment.
- Change the starting point: every close-out must state one judgment that doesn't depend on this round's materials.
- Write the decision into the judgment statement, and make the experiment owner accountable for it.
- From now on, sign off against one question: would this judgment still hold for a different product?
01How experiment records get translated
Putting experiment records in a document doesn't mean they've reached the close-out. The first step of translation is that every close-out states one judgment that doesn't depend on this round's materials.
If translation merely condenses the good-or-bad verdict, frontline teams still face the same problem: the next experiment can't borrow this one's judgment.
Every close-out must state one judgment that doesn't depend on this round's materials. Without that, experiment records remain stuck at whether the results were good or bad.
02Where experiment records get distorted
Experiment records are most often distorted when work gets split up. Each department takes a line that's easy to say, but not the refusal that goes with it.
So externally you still see close-outs that only record good or bad results, while internally no one can explain why the next experiment can't borrow this one's judgment.
03The smallest unit of translation for experiment records
The smallest unit of experiment records isn't a full manual but a judgment that can be used on the spot. The judgment statement should be that unit.
With experiment records in hand, frontline teams no longer have to guess what the brand team would think.
04Experiment records in frontline hands
At close-out, spot-check one live case to see whether the judgment statement has slipped back into a simple good-or-bad verdict.
Rewriting isn't necessarily bad. What's bad is rewriting without anyone acknowledging that the experiment records have changed.
05Where judgment lives in experiment records
Judgment should stay with the people who use it. The experiment owner is accountable, and the field has the authority to say no based on the judgment statement.
End training with one question: would this judgment still hold for a different product?
Framework
The four layers of experiment records
From the non-negotiable to tools usable in the field, each layer answers just one question, so experiment records don't slide back into a simple good-or-bad verdict.
L1Judgment
This layer sets out what in the experiment records is non-negotiable.
L2What applies only to this round
This layer turns experiment records into checkable conditions.
L3Transferable
This layer defines what experiment records may change across contexts.
L4Archive
This layer hands experiment records to the people who use them, and documents the exceptions.
Note: This diagram is a conceptual framework. It explains a way of thinking and is not statistical data. Source: Qingmei Future Insights.
Recommended actions
Action 01
Start from the site
At close-out, mark the moment experiment records are actually called on. Don't start by rewriting the good-or-bad verdict.
Action 02
Keep it to one page
Replace longer explanations with the judgment statement, keeping only what supports “every close-out must state one judgment that doesn't depend on this round's materials.”
Action 03
Judge it by one question
At close-out, ask only: would this judgment still hold for a different product? If you can't answer, it isn't a decision yet.
Action 04
Name the owner
Owner: the experiment owner. Experiment records without a maintainer get quietly rewritten whenever conflicts arise.
Related
All insightsRelated insights
The Reason to Stay Is Different from the Reason to Buy Once
In customer engagement plans, using acquisition messages to retain existing customers often means people have already joined but still hear no reason to stay.
Service Recovery Is Marketing Too, Beyond the Remit of Customer Support
After a service failure, asking customer support only to soothe emotions often leaves the brand absent at the moment people remember most.