· Orqestra Team · 7 min read
Restaurant Sales by Daypart: How to Analyze and Act on It
How to track restaurant sales by daypart from Clover, Square or Toast reports, build a weekly table with labor hours, and use it for scheduling and hours.
Sales by daypart means splitting your sales into the blocks of the day you actually run: breakfast, lunch, afternoon, dinner and late night. You get the numbers from your POS by filtering a sales report to a time window, then you line each window up against the labor hours you scheduled in it. The result tells you which hours earn their keep, which hours you are overstaffing, and whether that early open or late close is worth it.
This guide defines the dayparts, shows where the hourly data lives in Clover, Square and Toast, and walks through a full week for a one-location restaurant doing about $80,000 a month.
Define your dayparts first
There is no industry-standard clock for dayparts. YipitData, which tracks card spending, defines late night as 9pm to 5am and found that about 11% of 2021 restaurant sales happened after dinner, a larger share than breakfast (yipitdata.com). Beyond that, you set the boundaries to match your own service.
A workable default for a restaurant open 7am to 11pm:
| Daypart | Hours | What usually drives it |
|---|---|---|
| Breakfast | 7:00 to 11:00 | Coffee, grab-and-go, regulars |
| Lunch | 11:00 to 14:00 | Peak transactions, lower ticket |
| Afternoon | 14:00 to 17:00 | Slowest block, prep and admin time |
| Dinner | 17:00 to 21:00 | Highest ticket, highest labor |
| Late night | 21:00 to close | Bar sales, delivery, small crew |
Two rules when you draw the lines. First, boundaries go where staffing changes, not where the menu changes. If your dinner crew clocks in at 4:30, that is where dinner starts for analysis. Second, keep the same boundaries for at least eight weeks so you can compare like with like.
Where to get hourly sales from your POS
You need sales in time blocks. Every major POS can do this, but the path differs.
Clover. The Sales report on the web dashboard has a Time range control. Clover's help page says you can "select the default business hours or available dayparts," or pick "Custom day part" and enter start and end times, which "let you view your data for a specific time window across your selected date range. For example, you can see sales only from 7:00 PM to 11:00 PM each day" (clover.com help: run or request Sales reports on the dashboard). Dayparts themselves are configured under business hours. You can export the report from Options > Export. There is no single report that shows every hour of every day side by side, so for five dayparts you run the report five times with different custom windows, or export the full order list and group it yourself. The full walkthrough of the report is in how to read a Clover sales report.
Square. In the Square Dashboard sales summary, Square's help article says you can "choose to summarize by group, hour, or day" and "set individual increments by hours, days, weeks, or months," then print or export the report (squareup.com help: sales summary, sales trends and payment methods reports). Summarize by hour, export, and sum the hours into your dayparts.
Toast. Toast's Sales Summary report has a "More filters" option to narrow results by "revenue center, service area, dining option, or hours of the day," and it downloads as CSV or XLS (support.toasttab.com: How do I access sales reports). Toast also has an Hourly Sales report under Labor that "shows a breakdown of sales in 15-minute intervals" with labor cost as a percentage of net sales at each interval (support.toasttab.com: Labor Reports Overview).
Whichever POS you use, pull net sales (after discounts and refunds, before tax and tips). Clover's definition is "total item sales after discounts and refunds," and that is the number to compare against labor.
Step by step: build the weekly daypart table
- Pick one full week, Monday through Sunday, that was not a holiday week.
- Export net sales for each daypart window from your POS using the methods above. Five windows, seven days, 35 cells.
- Count transactions per daypart too. You will want average ticket later.
- From your schedule or time clock, total the labor hours actually worked in each window. Include kitchen, counter and managers who were on the clock. Split a shift that crosses a boundary by the hours on each side.
- Multiply hours by your average loaded hourly cost (wage plus payroll taxes; a reasonable shortcut is wage × 1.1) to get labor dollars per daypart.
- Calculate the two ratios for each daypart:
Sales per labor hour (SPLH) = net sales ÷ labor hours
Labor cost % = labor dollars ÷ net sales × 100
- Repeat for four to eight weeks before you change anything. One week tells you very little.
Worked example: one week at an $80k-a-month restaurant
A fast-casual place, open 7am to 11pm, $18,500 in net sales for the week, 329 labor hours, average loaded cost $16 per hour.
| Daypart | Net sales | Transactions | Avg ticket | Labor hours | Labor $ | SPLH | Labor % |
|---|---|---|---|---|---|---|---|
| Breakfast (7 to 11) | $2,100 | 210 | $10.00 | 56 | $896 | $37.50 | 42.7% |
| Lunch (11 to 14) | $6,300 | 420 | $15.00 | 84 | $1,344 | $75.00 | 21.3% |
| Afternoon (14 to 17) | $1,400 | 100 | $14.00 | 49 | $784 | $28.57 | 56.0% |
| Dinner (17 to 21) | $7,500 | 340 | $22.06 | 105 | $1,680 | $71.43 | 22.4% |
| Late night (21 to 23) | $1,200 | 70 | $17.14 | 35 | $560 | $34.29 | 46.7% |
| Week total | $18,500 | 1,140 | $16.23 | 329 | $5,264 | $56.23 | 28.5% |
The week total looks healthy: 28.5% labor is inside the 28% to 32% range Netchex gives as a fast-casual target (netchex.com). The dayparts tell a different story. Lunch and dinner carry the restaurant. Breakfast, afternoon and late night together take 140 of 329 labor hours (43%) and produce $4,700 of $18,500 in sales (25%).
Read the SPLH column against published ranges with care, because sources disagree. 7shifts lists $80 to $120 for fast casual and says "there's no universal 'good' SPLH" (7shifts.com); The Restaurant HQ lists $100 to $200 for fast casual and $60 to $100 for full-service casual (therestauranthq.com). The spread comes from different definitions of sales and hours. Your own eight-week history is the benchmark that matters; the published ranges only tell you whether you are in the neighborhood.
What to do with the table
Scheduling
The afternoon block is the obvious target. 49 hours for $1,400 in sales means seven people on the clock for three hours producing under $30 an hour each. Cutting that to 28 hours (four people) saves $336 a week, about $1,450 a month, and moves afternoon labor from 56% to 32%. Most of those hours were prep, so check whether prep can shift into the last hour of lunch when the crew is already there.
Set an SPLH target per daypart, not one target for the day. 7shifts makes the same point: "your lunch SPLH target will differ from your dinner target." For this restaurant, $70 at lunch and dinner is proven; $45 might be a fair stretch goal for breakfast and late night.
Promotions
Promote into weak dayparts that already have the right traffic, not into dead ones. Breakfast has 210 transactions at a $10 ticket. A $3 add-on pastry bundle that lifts breakfast ticket to $12 adds $420 a week without a single extra labor hour. Late night, with 70 transactions, needs new guests, not a bigger ticket, which is a harder and more expensive problem. More on ticket math in average ticket size for a restaurant.
Hours of operation
Late night costs $560 in labor to make $1,200. Whether that is worth keeping depends on the contribution after food cost. At a 30% food cost, late night contributes $1,200 − $360 − $560 = $280 a week before rent and utilities, which are paid either way. That is a thin but positive number. If the late crew also closes and cleans, closing at 9 would not remove all 35 hours; perhaps 20. Rerun the math with the hours you would actually save before you cut a daypart.
Opening earlier works the same way in reverse. Only extend hours when a neighboring daypart's SPLH is strong enough to suggest unmet demand, for example a lunch that is at capacity from 11:00.
Tracking it every week
Pulling five daypart windows from a POS report by hand takes about 20 minutes a week once you have the routine. Put the results in one sheet with a row per week and the eight columns above, and add a 4-week average so one odd week does not drive decisions. If you want this automated, the sales analytics feature breaks Clover sales (Square and Toast in early access) into hourly and daypart views without the manual export.
For what counts as a healthy overall labor number by restaurant type, see restaurant labor cost percentage by type.
Common mistakes
- Using gross sales or amount collected instead of net sales. Tips and tax inflate the numerator and make every daypart look better than it is.
- Moving daypart boundaries week to week. Pick them, write them down, leave them for two months.
- Assigning a whole shift to one daypart. A 3pm to 11pm server covers afternoon, dinner and late night; split the hours.
- Judging dayparts on labor % alone. A 45% breakfast with 200 transactions may be a growth daypart; a 45% late night with 70 transactions may be a closing decision.
- Acting on one week. Weather, a local event or a sick cook can swing a daypart by half. Use four to eight weeks.
- Cutting the daypart instead of the hours. The fixed costs stay; only the marginal labor and the marginal sales change. Run the contribution math first.
- Comparing your SPLH to a blog benchmark without checking how it was defined. Net or gross sales, with or without managers, makes a $30 an hour difference.