Your final recommendation should follow the evidence, even when it contradicts the hypothesis you started with.

A hypothesis is useful because it focuses the investigation. It is not something you need to defend after the data points elsewhere.

You can briefly acknowledge the shift if it helps explain the logic: "We initially suspected demand, but the analysis shows demand is stable and the main issue is cost inflation." That demonstrates learning rather than inconsistency.

Lead with the final answer, supporting evidence, risk, and next step. Mention the changed hypothesis only when it clarifies how the case evolved; do not turn the conclusion into a recap of every branch you tested. That is usually a stronger signal of judgment than staying attached to the opening hypothesis after the facts have changed. For more detail, see the case interview guide.