To find out how much you spend on eating out, take three months of bank and card statements, tag every restaurant, takeout, coffee, delivery and bar charge, and add them up. Divide by the number of weeks covered and multiply by 52 for a yearly figure. Most people find the real number is higher than their guess.
Which statement rows count as eating out?
Eating out covers any food or drink someone else prepared for you: sit-down restaurants, takeout, fast food, coffee shops, bakeries, delivery apps, bars, and food bought at work canteens or stadiums. Groceries do not count, even ready meals from the supermarket, because they sit in a different budget line.
Decide the boundary before you start counting, and write it at the top of the sheet. The arguments people have with themselves are predictable. Is a $5 coffee eating out? Yes, if you want an honest total, because coffee is often one of the largest groups by count. Is a pizza ordered through a delivery app eating out? Yes, and the delivery and service fees belong with it. Is a sandwich from the supermarket's hot counter? That one is a judgment call; pick a rule and keep it.
Keep groceries in view as the comparison. The point of counting eating out is usually to decide whether to shift some of it to cooking, and that only makes sense next to what you spend at the store. Comparing your grocery bill with the average from your statements does the same count on the other side of the food budget.
Sub-groups worth keeping
Five sub-groups make the total easier to act on:
- Sit-down restaurants
- Takeout and counter food
- Coffee and bakery
- Delivery apps
- Bars and drinks out
Each has a different fix. A family dinner out once a week is a choice; three delivery orders a week is often a habit.
Tagging restaurant and takeout rows in a spreadsheet
Tagging restaurant rows means giving each food charge one of the five sub-group labels in a spreadsheet column. Sort the sheet by description so repeat places line up together, tag the obvious ones in blocks, then scan the rest. Most of the work is the first pass; later months reuse the same names.
The statement description does not always say "restaurant". Small cafés often bill through a card processor and appear with its prefix in front of their name. Restaurant groups bill under a company name. Delivery orders usually show the app's name, not the restaurant's. When a description stumps you, check the amount against your email receipts for that day. If it still means nothing, leave it untagged; a stray $9 row will not change the conclusion.
Getting the rows into a sheet is the slow part when statements arrive only as PDFs. CheapBankConvert, which is getting ready to open, converts a PDF statement into CSV or Excel with the description and amount in separate columns, and the first 2 pages are free to preview. At early pricing a statement up to 10 pages is $2.99, or 5 statements for $9.99 in a bundle. The CheapBankConvert homepage has the current figures. If you compare other converters that sell prepaid packs for a long stretch of statements, read whether converter credits expire before you buy more than you need.
A formula that does the first pass
If the same places come back month after month, a small lookup list saves retagging. Put merchant names in one column of a second tab and their sub-group in the next. A lookup formula on the main sheet, matching part of each description against that list, fills most tags automatically. Anything it misses, you tag by hand and add to the list.
How much do I spend on eating out per month and per year?
To turn tagged rows into totals, add up each sub-group for the period, divide by the number of weeks it covers, and multiply by 52 for a year or by 52 and divide by 12 for an average month. Using weeks, not statement months, keeps the arithmetic honest when statement periods do not match calendar months.
A pivot table does this in seconds. Put the sub-group label in rows and the sum of amounts in values. Add a count of rows next to the sum, because the number of visits tells you as much as the money. Fifteen takeout meals in twelve weeks is a different habit from fifteen in four.
Three months is the minimum for a fair count. One month can be distorted by a birthday dinner, a holiday or a week of illness when nobody cooked. Twelve weeks smooths most of that out. If you went on a trip during the period, tag those meals separately as travel so they do not inflate your everyday figure.
The per-visit number
Divide each sub-group's total by its count to get the average spend per visit. That figure is what you change when you trim: a cheaper coffee order, a lunch without a drink, a delivery order collected instead. Knowing the average makes a small change easy to price.
Eating out by the numbers: one person's twelve weeks
The example below is one person's eating-out rows from twelve weeks of checking and card statements, grouped and totalled. The amounts are made up for illustration. Yearly figures are the twelve-week total times 52 divided by 12.
| Sub-group | Visits in 12 weeks | Total in 12 weeks | Per average month | Per year |
|---|---|---|---|---|
| Sit-down restaurants | 7 | $412.60 | $148.99 | $1,787.93 |
| Takeout and counter food | 15 | $238.50 | $86.12 | $1,033.50 |
| Coffee and bakery | 34 | $173.40 | $62.62 | $751.40 |
| Delivery apps | 6 | $201.30 | $72.69 | $872.30 |
| Bars and drinks out | 4 | $96.00 | $34.67 | $416.00 |
| Total | 66 | $1,121.80 | $405.09 | $4,861.13 |
This person guessed they spent "about $200 a month" on food out. The statements say just over $400, roughly twice the guess. The gap is typical, because the small, frequent purchases are the ones memory drops. Coffee alone is 34 visits, more than half of all visits, yet only about 15 percent of the money.
The delivery line tells a different story. Six orders cost $201.30, an average of $33.55 each, against $15.90 for a takeout meal collected in person. Some of that gap is bigger orders, but a share of it is fees and higher menu prices. For a closer look at that markup, estimating food delivery fees from one statement charge breaks a single order into its parts.
What this person might change
Looking at the table, the options with the most money in them are clear. Swapping half the delivery orders for collected takeout would save about $230 a year at the example's averages. Dropping one sit-down dinner a month would save more. Cutting coffee would save the least per decision, though it might be the easiest habit to change.
How does your eating-out total compare with the average household?
US households spent an average of $3,945 on food away from home in 2024, against $6,224 on food at home, according to the Bureau of Labor Statistics' consumer spending summary. That is about $329 a month for eating out. Compare your yearly total with that figure, keeping in mind it is an average across all household sizes.
The comparison needs care. The BLS figure is per household, and households range from one person to large families. A single person matching the household average is spending more per head than a family of four doing the same. An average also hides big differences by income and region. Use it as a rough reference point, not a target.
Your own ratio is often more useful than the national number. In the BLS figures, eating out is a little under 40 percent of total food spending. If your eating-out total is larger than your grocery total, that is a clearer sign of where the money goes than any average. The person in the worked example would need their grocery bill alongside to make that check.
When a high number is fine
Spending more than average on eating out is not automatically a problem. If you cook little because you work long hours, travel for work or live alone in a city with cheap food, a higher figure may be a fair trade. The count matters because it lets you decide on purpose, rather than discover the number by accident.
Eating-out edge cases: split bills, work lunches and gift cards
A few situations make the eating-out total from statements wrong unless you adjust for them: shared bills paid on one card, meals your employer will reimburse, and food paid with gift cards or cash. Each one moves money in or out of the count without changing what you personally spent.
Split bills. If you paid for a table of four and three friends sent you their share by a payment app, the statement shows the full bill as your spending and the repayments as separate incoming rows. Net them against the meal, or your total will be inflated. The same applies in reverse when a friend paid and you reimbursed them; that payment is eating out even though no restaurant name appears.
Work meals. Meals your employer reimburses are not your spending once the money comes back. Tag them separately and leave them out of the total, but check that the reimbursement actually arrived.
Gift cards and cash. A meal paid with a gift card or cash does not appear on any statement. If you use cash for coffee or lunch often, the statement total is a floor, not the real figure. Keep a rough note of cash food spending for the same twelve weeks.
Unplanned treats. Some food rows are impulse purchases rather than meals: a snack at a gas station, a pastry on the way past. If you want to see how many of those there are, and when they happen, finding impulse purchases on bank statement rows shows how to filter for small, unplanned charges by time and place.