A cover is a single paying guest. A reservation for four equals four covers, not one ticket.
Industry usage: "We did 220 covers Saturday" = 220 paying guests served. "Average check is $24 across 1,650 weekly covers" = $39,600 in weekly sales.
Why the unit matters
Operators who report "covers" interchangeably with "tickets" or "tables" produce metrics that don't compare across reporting periods. A four-top is one ticket, one table, and four covers — three different denominators. Every operational metric that uses a per-guest denominator (average check, sales per cover, covers per labor hour) needs the actual paying-guest count, not the table or ticket count.
The standard:
- Cover: one paying guest. Children at the table count as covers if they ordered (even just a kids' meal). Comp guests count as covers — they were served, even if the revenue was zero.
- Ticket: one check. A four-top with a single ticket is one ticket and four covers. A four-top with two split checks is two tickets and four covers.
- Table: one party. A four-top is one table and four covers.
How to count covers correctly
- Pull the per-guest count from the POS, not the ticket count. Most modern POS systems track guests per ticket as a separate field — make sure servers are entering it.
- Decide your comp policy upfront. Either count comp guests as covers (industry standard) or don't — be consistent across reporting periods.
- For takeout / delivery: 1 ticket = 1 cover by convention, even if multiple meals were ordered. The metric breaks down for delivery anyway; don't try to back-calculate "household covers."
Common operator mistakes
- Reporting tickets as covers. Inflates per-guest spend and misstates capacity. A $96 four-top reads as a $96 average check if you divide by 1 ticket and a $24 average check if you divide by 4 covers — same revenue, totally different signal.
- Excluding comp guests from cover counts. Distorts both average check and covers-per-labor-hour. Industry convention: comp guests count.
- Inconsistency across servers on guest-count entry. If half the team enters guests-per-ticket and half doesn't, the data is unusable. Make it a closing-shift requirement.
Related concepts