The Focus Group That Disagreed With the Data
A regional bank is deciding whether to launch a fee-free checking account aimed at 25–34 year-olds. A 90-minute focus group of 8 participants in one metro area reacted enthusiastically; several said the no-fee positioning was 'refreshing' and 'the reason I'd switch.' Based on this, the product team wants to greenlight a full launch. However, a separate online survey of 1,200 respondents in the target age band found that only 22% ranked fees as a top-three switching driver, behind mobile app quality (61%) and branch/ATM access (44%). The survey sample skewed slightly toward higher-income respondents, and the focus group was recruited from a single urban market. The CMO is skeptical of the survey and wants to trust the qualitative signal.
1. Evaluate the strength of each evidence source and explain why they diverge.
Read the worked answer
The focus group offers depth but low generalizability: 8 people in one metro, subject to social desirability and groupthink, and enthusiasm in a moderated setting often overstates real switching behavior. It is useful for generating hypotheses and language, not for sizing demand. The survey offers breadth and quantifiability (n=1,200) but has its own biases: self-reported importance rankings can understate price sensitivity, and the higher-income skew may depress the salience of fees relative to the broader target. The divergence is expected, not contradictory - qualitative enthusiasm reflects emotional resonance with the message, while the survey measures stated importance of attributes, which is a different construct. Neither alone is decision-grade, and the survey's 'top-three driver' framing may miss that fees can be a threshold/hygiene factor that rarely ranks first but still blocks switching.
2. Recommend how the bank should resolve the conflict before committing to a full launch, including the specific research design.
Read the worked answer
Triangulate rather than pick a side. Run a conjoint or MaxDiff study on a representative sample of the target to quantify trade-offs among fees, app quality, and access - this forces respondents to make choices instead of rating importance in isolation. Weight or re-weight the sample to match the target demographic, and check whether fee sensitivity differs by income segment. Complement with behavioral evidence: a limited-market pilot or A/B test of the offer with real signup and activation data, since stated preference diverges from behavior. Use the focus group only to refine messaging and identify objection language. Decision rule: proceed to full launch only if the conjoint shows fee-free positioning meaningfully shifts preference share and the pilot shows acceptable activation and early attrition.
3. The CMO says 'people don't know what they want; trust the room.' How would you respond, and what would change your mind toward the CMO's view?
Read the worked answer
The CMO is right that stated importance can mislead - customers often can't articulate trade-offs, and a single survey question is weak evidence. But 'trust the room' overgeneralizes from 8 people in one market and confuses resonance with demand. The disciplined response: treat both as inputs and let behavioral data arbitrate. I would change my view toward the CMO if the conjoint showed fees dominating preference share despite the survey, or if a pilot showed fee-free messaging driving materially higher qualified signups and retention than the survey implied. If behavioral evidence consistently contradicts the survey, the survey instrument - not the qualitative signal - was the problem.