Restaurants & Food · Scenario 02

Specialty Coffee Café

Mornings are a wall of people; afternoons are a ghost town. The café loses sales to the line at 8am and money to idle staff at 3pm. We treat the counter as a queue and re-engineer it.

Method · M/M/c queue modelling

The situation

At peak the line spills out the door and a measurable share of would-be customers glance at the queue and keep walking — revenue that never even rings up. The owner’s instinct is to add a barista to every shift, but labour is the single biggest cost and most of the day the café is over-staffed.

Nobody has measured the actual arrival pattern, the service time per drink, or the point at which the line gets long enough to drive people away. Staffing is set by habit and goodwill, not by the rhythm of demand.

~9 min

peak wait at the till

18%

walk-aways at peak

1 rota

same staffing all day

42%

idle barista hours off-peak

Where we dig for the truth

We time the café like an operations problem: when people arrive, how long each drink takes, and how the line behaves as both change through the day.

POS timestamps (per order)Drink-level service timesDoor / foot-traffic countsStaff rota & labour costAbandonment observationsWeather & daypart

An M/M/c queue model shows wait time explodes past a certain demand-to-barista ratio — but only for a 90-minute window, not all day.

Our approach — Queueing-Theory Staffing

Using arrival rates and service times we model the counter as a multi-server queue. That tells us precisely how many baristas each 30-minute block needs to keep the wait below the threshold where customers walk — no more, no less.

The fix is rarely ‘more staff’. It is the right staff in the right 30 minutes, a faster path for the five drinks that make up most of the volume, and mobile pre-orders to flatten the morning spike.

Peak wait drops from nine minutes to three with no net increase in labour — staff simply move from idle afternoons to the morning rush.

What changes

Same team size, re-timed to demand, plus a faster line. Representative for a café doing roughly 700 transactions a day.

Why this is not "social media management"

We didn't grow this café with a louder Instagram. We grew it by counting seconds and people, then spending the exact labour the morning needs and not a dollar more. The queue was the marketing problem — and queues are solved with math, not hashtags.

Frequently asked questions

How do you reduce café wait times without hiring more staff?

We model the counter as a multi-server queue (M/M/c) from your POS timestamps, then match barista hours to each 30-minute block of demand. Staff shift from idle afternoons to the morning rush, cutting peak waits from around nine minutes to three with no net increase in labour.

What is queueing theory?

Queueing theory is the mathematics of waiting lines — how arrival rates and service times determine how long customers wait and when a line gets long enough that people leave. It tells us the exact number of servers needed to keep waits below the walk-away threshold.

How is this different from a typical marketing agency?

Instead of ads or social posts, we fix the operational bottleneck quietly costing sales — the line at the till. It is statistics applied to throughput, not a campaign. Book a marketing audit.

Want this run on your numbers?

Share a week of POS timestamps and we’ll model where your line is costing you customers.