Turning scattered data into decisions leadership trusted.

Sales lived in the POS, labour lived in scheduling, and customer and campaign data lived in marketing tools. Each team pulled its own export, which meant weekly reviews often opened with a debate about whose number was right.

The cost wasn't just time. When the numbers were contested, decisions got deferred — and deferring a labour or menu decision for a week is expensive across nine locations.

I started with the questions leadership actually asked each week, then worked backwards to the smallest set of metrics that answered them. Anything that didn't change a decision didn't make the dashboard.

Getting agreement on definitions came first — what counts as a transaction, how labour hours map to a daypart, which channel gets credit for a delivery order. Once those were written down, the pipeline was straightforward: consistent extracts from ParPOS, labour, and web analytics feeding a shared reporting layer.

I built the dashboards to be read in two minutes, with drill-down available for anyone who wanted the detail behind a number.

Weekly reviews shortened because the meeting started from a shared view instead of reconciling spreadsheets. Operations and marketing began quoting the same figures.

Store-level performance became comparable, so labour and menu decisions were made against evidence rather than intuition — and the leadership team could see the effect of a change the following week.

The first version shipped with more charts than anyone needed, which diluted attention. Launching with a handful of decision-driving metrics and adding on request would have driven adoption faster.