Comparison · C92
Assess the evidence level of a marketing initiative
What evidence is enough to test, fund, scale or stop an initiative?
Direct answer
Opinion, analogy, observation, quasi-experiment and controlled experiment support neither the same commitment nor the same claim.
01Cap investment according to evidence strength and the cost of error.
02The rule, the evidence and the next review date stay visible.
How to use this
Move from the question to the deliverable.
What evidence is enough to test, fund, scale or stop an initiative?
Cap investment according to evidence strength and the cost of error.
Critical comparison of five evidence levels
Critical comparison · SQ-92
Match the commitment to the strength of the evidence
Weak evidence can authorise exploration, never an irreversible rollout. The evidence method is chosen according to the decision and the cost of an error.
Comparison applied to a fictitious ABM initiative whose funding could move from €15k to €320k.
- 01Name the claim
State what the evidence must show: need, access, conversion, causality or economics.
- 02Classify
Identify the level actually reached, without turning a signal into a demonstration.
- 03Cap
Set a maximum financial and operational commitment for each level.
- 04Strengthen
Choose the next piece of evidence that most reduces the uncertainty that matters for the decision.
Five levels of evidence, five commitment limits
Causal strength rises, but so do the time and cost involved. The best evidence is whatever is enough for the next decision.
| Level | Example | Causality | Duration | Main bias | Authorised commitment |
|---|---|---|---|---|---|
| Expert opinion | Internal interview | 1/5 | 1–3 days | Confirmation | Exploration ≤ €5k |
| Analogy | Comparable case | 1/5 | 1 week | Transferability | Design ≤ €15k |
| Observation | Exposed cohort | 2/5 | 2–6 weeks | Selection | Pilot ≤ €50k |
| Quasi-experiment | Comparable group | 4/5 | 6–12 weeks | Residual confounding | Expansion ≤ €150k |
| Controlled experiment | Random assignment | 5/5 | 8–16 weeks | External validity | Scale ≤ €320k |
01The interviews authorise designing the pilot, not forecasting revenue.
02A quasi-experiment is enough to extend to two segments when the cost of waiting exceeds the residual risk.
03The controlled experiment is reserved for scaling, because the time it takes is not justified before the pilot.
Methodological comparison. The ceilings are fictitious and must reflect the actual irreversibility and cost of error.
Coherence controls
Minimum conditions before validation.
The status, the evidence and the owner must be available for review.
The status, the evidence and the owner must be available for review.
The status, the evidence and the owner must be available for review.
The status, the evidence and the owner must be available for review.
Usable deliverable
Critical comparison of five evidence levels
Explore the decision further with an interactive control, or start from a structured template to produce your deliverable.