Ask what the data can answer.

The Kopi & Co exercise began with a very definite brief: a fictional COO, forty minutes, three decisions and exactly three screens. I liked having something concrete to build. There was also a question to ask before filling each part of the interface: did the records contain what that decision would require?¶

In my early individual branch, one row described an outlet over a week. That made comparisons between outlets possible. If waste looked unusual, I could identify somewhere worth investigating. A product decision needed another level of detail: which products, with what sales, waste and margins?¶

Labour raised a similar question. Weekly hours and an efficiency gap could help begin a discussion. To build a workable roster, I would also need to know when demand arrived, who was available, which skills were required and what service coverage had to be maintained. The next useful question became much more specific.¶

At the 7 September presentation, I explained why outlet-level waste could not establish which product line to pull. Our team had built a fuller demonstration using synthetic product records. My separate branch kept its original data scope. Looking back, I want to keep the purpose and provenance of those versions easy to follow.¶

For my October individual reflection, I returned to the experience using framing and reframing. Prioritising an investigation, specifying the evidence for an intervention, and demonstrating a future workflow are different tasks. Writing those alternatives down helped me see what a satisfactory answer would look like in each case.¶

I also remembered the smaller audit questions. A complaint percentage could look dramatic because its starting rate was low. A priority change could leave the shortlist unchanged if no outlet triggered the relevant rule. Asking why the display looked that way was part of understanding the result.¶

If I began the sprint again, I would write down the three decisions and the evidence each needs before building. Then I would separate the useful next action from the questions that still need an answer. Sometimes a shorter, more precise list gives the following conversation a better place to begin.¶