What 6 months of queue data taught us about retail throughput

Six months ago we wrapped a queue-measurement deployment across twelve sites. Here's what the data said.

The hypothesis

Going in: more checkouts = shorter queues.

What the data said

No. Throughput was bottlenecked not at checkout but at the back of the queue. Customers were abandoning before they reached a register.

What worked? Signage with live wait-time estimates derived from upstream counts. A 30% drop in abandonment within four weeks.