The lowest labor for this time of year in nine years
In its first seven weeks on Shiftova, this restaurant's labor came to 21.0% of sales, lower than the same seven weeks of every year back to 2017. Labor was already coming down before the switch, so we can't say how much of that is Shiftova. What we can measure is the sales forecast the schedule was built on.
Sep 13–Oct 31, each year
Labor as a share of sales
2026: 21.0%, against 24.3% for the same weeks last year. 2021 isn't shown: its September records are damaged.
- The restaurant
- Single-location pizza, British Columbia
- The period
- First 7 weeks on Shiftova, Sep 13–Oct 31, 2026
- Uses Shiftova for
- Schedules, time clock, payroll hours, sales forecast
- Compared with
- The restaurant's own records, back to 2017
This restaurant is owned by Shiftova's founder. It's where every feature gets tested first.
21% of sales, down from 24% last fall
Labor here is the cost of staff hours as a share of net sales, from the restaurant's monthly reports. Across its first seven weeks on Shiftova it came to 21.0%. The same weeks of every other year on record ran between 23.5% and 27.1%.
Two things to know before giving the app the credit. Labor drops here every fall: sales pick up after summer while the crew stays about the same size, and last year it fell 4.4 points between summer and fall. And this year was already running leaner than last, with about 6% fewer staff hours at the same sales from January to mid-September, before Shiftova was on the schedule.
A forecast that missed by less, with half as many bad days
Every schedule here starts from a sales forecast for each day. Before Shiftova, that was the average of the same weekday over four recent weeks. We replayed both for each of the 49 days, using only the sales known when that week's schedule was built, and compared them with what actually sold.
- Off for every $1,000 of sales
- $207$167
- Off per day, in staff-hours
- 5½ hrs4½ hrs
- Days off by more than 30%
- 137
- Thanksgiving weekend, Sat–Tue
- 51% over15% over
- Halloween, Fri + Sat
- 33% short3% short
Crossed out: the old four-week average. In bold: Shiftova.
Within 10%
10–20% off
20–30% off
More than 30% off
Days, out of 49
On ordinary days the two were close, and the old average actually landed within 10% on two more days. Shiftova's edge is the big misses, the days a schedule goes most wrong.
Show the numbers as a table
| How far off | Old four-week average | Shiftova |
|---|---|---|
| Within 10% | 21 days | 19 days |
| 10–20% off | 11 days | 13 days |
| 20–30% off | 4 days | 10 days |
| More than 30% off | 13 days | 7 days |
Where the old average fell apart
Thanksgiving weekend
Thanksgiving weekend runs slow here every year. Built from four normal weeks, the old average had the Sunday at more than double what sold, and the four days together 51% over. Shiftova had the four days 15% over, and was closer on every one of them.
Read the Thanksgiving case study →Halloween on a Saturday
Part of the rush moved to Friday. Shiftova had Friday 29% too low and Saturday 22% too high, but the two days together within 3%. The old average was 33% short over the two days, with Saturday at about half of what sold.
Where it missed
Shiftova's biggest miss was the Wednesday after Thanksgiving, forecast 80% too high (the old average was 59% too high). It also had Truth and Reconciliation Day 38% too high, where the old average was 18% high. Those are the days to adjust yourself on the schedule, the same way you would for a catering order or an early close.
Questions owners ask
Did Shiftova lower the restaurant's labor?
Not something we can prove yet. Labor drops here every fall, and this year was already running leaner than last year before the switch. What we can measure is the forecast the schedule was built on, so that's what this page credits to Shiftova. We'll look at labor again once November and December are in.
What was the restaurant using before?
The average of the same weekday over four recent weeks, a common rule of thumb. Because a schedule is built a week ahead, it uses the four weeks ending the week before last.
How much sales history do I need?
Forecasts start after four weeks of daily sales. Shiftova picks up the time of year once it has a full year, and a holiday's own pattern once it has seen that holiday two or three times. It gets sharper the longer you use it.
Do I need to connect my point-of-sale?
No. You type in each day's total sales. It takes a few seconds a day, or a minute a week.
What else comes with it?
The forecast is one part of Shiftova. The same app handles clock-ins with GPS, schedules and shift swaps, payroll hours with overtime and stat pay, and team chat. Sales forecasting and Draft week come with the Pro and Premium plans, and every plan starts with 3 months free with everything unlocked.
See what your own numbers say.
Start entering your daily sales and Shiftova starts forecasting in four weeks. 3 months free on any plan, cancel anytime.
Available on iOS and Android