Contribution by daypart and channel

A restaurant is not one business. It is usually three or four running through the same kitchen: a lunch business, a dinner business, a delivery business and, in many venues, a weekend business that behaves nothing like the weekdays. Managed as a single P&L, they hide each other.

## Why the total misleads

Revenue aggregates. Contribution does not. Two dayparts with identical revenue can have completely different contribution because the mix, the labour cost and the channel are different.

A lunch service running at high volume on low-priced, low-contribution items with full staffing can consume more than it earns while showing healthy revenue. A quiet weekday dinner selling few but high-contribution items can be doing better than it looks. The total tells you neither.

## How to split it

You need three things: sales by item, cost by item, and a timestamp or channel marker on each transaction. Most point-of-sale systems hold all three; few operators export them together.

**By daypart.** Split the day at the natural service boundaries rather than at the clock. If lunch runs 12:00 to 16:00 and dinner from 18:00, the gap belongs to whichever service prepares during it.

**By channel.** Dine-in, takeaway, own delivery, each platform. Platforms should be separate, not grouped, because commission terms differ.

**By day type.** Weekday and weekend at minimum. Many venues discover that their weekday business is being subsidised by two days a week.

Then, for each segment: revenue, food cost, and — where it can be attributed — the labour that segment actually requires.

## What usually shows up

Three findings recur often enough to be worth looking for specifically.

**A daypart that does not cover its labour.** Opening hours are usually inherited rather than decided. A service that runs at low volume with a full brigade is a fixed cost pretending to be a revenue stream.

**A channel that looks large and contributes little.** Delivery volume can be substantial while its contribution, after commission and packaging, is a fraction of the dine-in equivalent.

**A mix difference between dayparts.** Lunch and dinner often sell entirely different parts of the menu, which means a menu change that helps one can quietly damage the other.

## What to do with the finding

The instinct after this analysis is usually to close the weak daypart. That is occasionally right and often wrong, because some of the cost is fixed and will not disappear with the service.

The more useful questions are: can the labour be scheduled differently rather than removed; can the offer for that daypart be changed rather than withdrawn; and is the daypart earning something that does not appear in its own contribution, such as visibility or the habit that brings a guest back at a more profitable hour.

A quiet lunch that introduces guests who return for dinner is not a loss-maker; it is an acquisition channel. But that should be a demonstrated pattern, not an assumption used to protect a service nobody wants to close.

## Practical recommendations

1. Export sales by item with a timestamp and a channel marker for a full representative period.
2. Attach item costs and calculate contribution, not revenue, for each segment.
3. Attribute labour by schedule rather than by ratio — the rota tells you who was there.
4. Look for the segment with the widest gap between revenue rank and contribution rank. That is where the misunderstanding is.
5. Before closing anything, test whether its cost is genuinely variable.
6. Repeat the analysis quarterly. Mix drifts, commission changes and habits move.

## In short

Most operators know their busiest hours. Fewer know their most profitable ones, and the two are frequently not the same. Splitting contribution by daypart and channel is the fastest way to find out which part of the business is carrying the rest.