Cross-functional scenarios
Describe a situation where you had to make a marketing decision with incomplete data. What was your approach?
Describe a decision with a real deadline, where waiting for better data would have cost more than being wrong. Say what you knew, what was missing, and the assumption you chose and why.
Then show how you made the call small or reversible, the result that would have changed your mind, and what happened. Calibrated confidence scores higher than a lucky call told with certainty.
5 min read
Free 10-minute mock
Practice “Deciding with incomplete data” free
Your first mock interview is free (10 minutes). No card needed.
What the incomplete-data question is about
This is a question about judgment under a deadline. Product marketing rarely gets clean data, and waiting for it is often the more expensive mistake. Panels want to hear you reason in the open. What did you know? What did you assume? How did you keep the downside small?
On the rubric, Strong analytical rigor means saying plainly what the evidence does and doesn’t support. Exceptional means looking for the data that would disprove your own view and making the uncertainty visible. The most useful sentence to prepare is the one that starts “I would have changed my mind if.” Few people say it unprompted, and it lands.
Sorting what you knew from what you guessed
Before telling the story, sort your inputs into four piles. It keeps the answer honest and gives the panel a structure without announcing one.
- Facts. Things you could check, such as survey counts, usage data, or a deadline.
- Signals. Directional evidence, such as support tickets or sales anecdotes, that points one way without proving it.
- The assumption. The judgment you made to fill the gap, stated as a sentence you could be wrong about.
- The cost of waiting. What a later decision would have cost, which is what justified deciding at all.
People tend to describe the data they had in loving detail and never say the assumption out loud. That is the part being scored.
Reasoning through a split survey, out loud
“We were adding paid live tutoring to a language-learning app, and I had two weeks to pick the launch message. Our survey had forty responses, split almost evenly between people who wanted to speak confidently and people preparing for a language exam. That was the data. What I didn’t know was which group would pay. I assumed the exam group would, because they had a deadline and their support tickets already asked for human help. I led with exam readiness for the paid tier. To keep the risk small, I ran both messages in the in-app upsell for a week, split evenly. Before it started, I told my manager that if the confidence message won trial starts by more than two points, we’d switch. The exam message won trial starts. The confidence message drove more second sessions, though, which surprised me. So we kept exam readiness for the upsell and moved the confidence message into onboarding after someone booked. The part I’d stress is that one week in one channel told us about that channel, not about the whole market.”
The reasoning is visible at every step. The assumption is named with its reason, the switch rule was set before the test, and the surprise result changed the plan instead of being explained away. The last line limits the claim, which is what makes the speaker sound calibrated.
Conclusions an incomplete-data story shouldn’t claim
- That the test proved the assumption in general. A short test in one channel says something about that channel and that week.
- That instinct was vindicated. Even if the call was right, the story is stronger when the process would have caught a wrong call.
- That a small survey represents the market. Forty answers can show a split. They can’t size it.
Naming a limit out loud doesn’t weaken the story. It moves it from Developing, where data is decoration, toward Strong, where the evidence and the claim match.
Where incomplete-data answers lose points
The gut call told as courage
“We didn’t have time for proper research, so I went with my experience. I’d launched similar products before and had a good sense of what would resonate. The launch performed well, and I think sometimes you just have to be decisive.”
Experience is a real input, and this answer will sound fine to many people. It loses points because nothing was uncertain in the telling, nothing limited the risk, and nothing would have shown the call was wrong. Keep the experience, then add the assumption it produced and the check you ran.
Waiting, framed as rigor
A story about waiting for complete data can sound careful. Panels often hear slow.
Follow-ups on deciding with incomplete data
| Follow-up | What sits behind it | A strong reply |
|---|---|---|
| What data would have changed your decision? | Whether you held the view loosely | Give the specific result, ideally one you named before the test |
| Why were forty responses enough? | Whether you know what a small sample can show | Say it was enough to see a split and too small to size it, which is why you tested |
| What if the test had gone the other way? | Whether the switch rule was real | Describe exactly what you would have changed and how fast |
| How do you tell stakeholders you’re unsure? | Whether uncertainty costs you credibility | Describe stating the assumption and the review date together, so doubt comes with a plan |
More questions in this category
Frequently asked questions
- Is it fine to say I relied on intuition?
- Yes, as one input. Pair it with the assumption it led to and the way you checked it. Intuition with no check reads as luck, even when the result was good.
- What counts as incomplete data for this question?
- A small survey, conflicting signals from sales and product, or no data on a new segment all qualify. Pick one where the missing piece mattered to the decision.
- Does the decision need to have turned out right?
- No. A wrong call you caught quickly because you had set a check in advance can score higher than a right call you can’t explain.
Guides that help with this question
All guidesFree 10-minute mock
Practice “Deciding with incomplete data” free
A Mokk session asks this question, follows up on your answer, and scores it on five competencies. Your first mock is 10 minutes, scored like a full interview. No card needed.
Your first mock interview is free (10 minutes). No card needed.
Practice “Deciding with incomplete data” free
10 minutes · no card needed