What does this campaign comparison actually show?
Read three response measures without turning them into an invented sales result.
The question
Alder randomly assigns 400 adult renters considering storage to one exposure each: 200 see A and 200 see B. Before the study, correct association with the no-drilling promise is named the primary outcome. Recall and stated intent are secondary. All data are invented.
| Concept A | Concept B | |
|---|---|---|
| Participants | 200 | 200 |
| Unaided brand recall | 80 | 74 |
| Correct no-drilling promise association | 70 | 118 |
| Stated purchase intent | 50 | 68 |
People can count in several rows. These are separate responses, not funnel stages. There is no unexposed control and no purchase observation.
B lifted sales by 24%. It wins every brand measure. Roll it out nationally.
- Target rentersSeveral people stating purchase intent still wanted to check price and dimensions. No order was placed in the study.
“What does this campaign comparison actually show?”
Not saved by this site. Your browser may restore text when you navigate back.
A strong answer, step by step
Open each step in turn. This is one strong answer, not the only one.
Step 1: Use the predeclared outcome
What would you do here? Open the move you’d make.
Select whichever metric favors the preferred concept.
Tempting, but it misses something. That changes the criterion after the fact.
Add the three response counts into one conversion rate.
Tempting, but it misses something. The rows overlap and are not stages.
Start with promise association and report the other measures.
The strong move. This follows the stated job.
See how a strong candidate says it
“Compare promise association first, then show the secondary trade-offs.”
Why it works: Do not change the success criterion after seeing results.
Step 2: Calculate the difference
What would you do here? Open the move you’d make.
Report the 24 percentage-point difference in this sample.
The strong move. This matches the counts and measured outcome.
Report 24% higher sales.
Tempting, but it misses something. Purchases were not observed.
Divide each count by all 400 participants.
Tempting, but it misses something. That uses the wrong exposure denominator.
See how a strong candidate says it
“A is 70/200 = 35%; B is 118/200 = 59%. B is 24 percentage points higher on promise association.”
Why it works: Use each concept’s exposure count. Percentage points are not a sales percentage.
Step 3: State the trade-off
What would you do here? Open the move you’d make.
Call B the winner on every brand measure.
Tempting, but it misses something. Recall is lower for B in this sample.
Prefer B for the stated job and disclose the other measures.
The strong move. This connects the decision to the actual evidence.
Switch to A without explaining why recall replaces the objective.
Tempting, but it misses something. A changed objective needs an explicit reason.
See how a strong candidate says it
“B has lower recall, 37% versus 40%, and higher stated intent, 34% versus 25%. Prefer it for the clarity-led pilot.”
Why it works: This mixed result is not an estimate of absolute advertising lift or sales lift.
Step 4: Choose the next evidence
What would you do here? Open the move you’d make.
Book the intent count as forecast orders.
Tempting, but it misses something. Intent is not an order.
Repeat liking questions until the headline is higher.
Tempting, but it misses something. That does not close the purchase-evidence gap.
Observe actual purchases before making a sales claim.
The strong move. This measures the business result the draft asserted.
See how a strong candidate says it
“Run a bounded purchase-outcome pilot with an appropriate comparison and clear price and availability.”
Why it works: The follow-up leaves purchase conditions unresolved. No significance or causal sales claim was established here.
The answer, in short
- Decision
- Take B into a bounded next pilot for the no-drilling message.
- Because
- It has higher promise association in this sample. Recall, intent and purchases are different outcomes.
- What would change it
- A new comparison weakens the clarity finding, or a purchase pilot fails the business objective.
Common traps
- Calling percentage points sales lift.
- Adding overlapping response counts.
Check your answer
- What was the outcome before results?
- What is each denominator?
- Which measure favors each concept?
- Which outcome was never observed?
Last reviewed Nov 10, 2026