The Data Said Cut It. The Analysis Said Keep It.

A regional QSR chain was conducting a routine menu rationalization review. Analytics flagged a CPG item as underperforming on standalone sales velocity. The recommendation from the data team was elimination. The item was weeks from being cut.

Before the decision was finalized, a deeper diagnostic was run on transaction-level data — not just item sales, but basket composition and attachment patterns across dayparts and customer segments. The question wasn't 'how does this item sell?' but 'what does this item do to the rest of the basket?'

The CPG item was driving incremental adult meal attachment at a rate that made it one of the highest-margin contributors in the category when viewed at the basket level. Standalone velocity was low because the item was never meant to be a standalone purchase — it was a companion item that elevated the check average of the customers most likely to return.

The elimination decision was reversed. The item was repositioned in the menu architecture to make its companion role explicit. Attachment rates improved further. The financial reporting that had flagged it for removal had been measuring the wrong thing.