The Deadline the Evidence Cannot Meet
A consultant is conducting a customer-satisfaction diagnostic for a 60-location retail chain. The agreed method is a stratified survey of 1,200 customers plus 25 store visits, to be completed over 6 weeks. In week 4, only 380 responses have arrived (a 19% response rate against a 60% target), and 9 store visits are done. The client's CEO has a board offsite in 10 days and wants 'the satisfaction findings' presented there. The consultant's preliminary data show a wide gap between urban and rural stores, but the rural sample is thin (n=41) and the store visits are concentrated in one region. The engagement letter promises 'robust, statistically reliable findings.' The client is a long-standing account and the consultant's largest.
1. What should the consultant present at the board offsite, and how should it be framed?
Read the worked answer
Present what the evidence supports, clearly labeled by confidence, and explicitly state what it does not yet support.
Reasoning:
- The engagement promised 'robust, statistically reliable findings.' The current data cannot deliver that: a 19% response rate raises non-response bias, the rural sample (n=41) is too thin to support a rural/urban conclusion, and visits are regionally concentrated.
- The urban/rural gap is a promising signal, not a finding. Presenting it as a finding to a board would overstate the evidence and could drive a national decision on thin data.
- The client's deadline is real; the consultant's job is to serve the meeting honestly, not to fill the slot with unsupported conclusions.
Framing: 'Here is what we can say with confidence (urban-store signal, based on n=X), here is what is directional but not yet reliable (the rural gap, n=41), and here is what we will know by [date] once fieldwork completes.' Recommend the board treat the directional items as hypotheses to test, not decisions to act on. Offer a follow-up briefing when data are complete.
2. The CEO says: 'Just give me the headline - urban stores are underperforming, right?' How should the consultant answer?
Read the worked answer
Give the honest headline: there is a signal, and it is not yet reliable enough to state as a conclusion.
Reasoning:
- Confirming 'urban stores are underperforming' would convert a thin, regionally skewed signal into a board-level assertion. If the board acts on it (reallocating investment, changing store formats), the cost of being wrong is high and the consultant owns it.
- The consultant should not hide behind 'more research needed' either. The client is entitled to the consultant's current best read, with its confidence level attached.
A sound answer: 'The early signal points that way - urban scores are running below rural in the data we have. But the rural sample is 41 responses from one region, so I can't yet tell you whether that's a real urban/rural pattern or a regional one. I'd give you a directional read now and a defensible answer in three weeks. If the board needs to act in 10 days, let's frame it as a hypothesis to test, not a conclusion.' This respects the deadline, the relationship, and the evidence.