Guide Polski

What you really spend on food — working it out from receipts

Ask anyone what they spend on food each month and you’ll get a figure that’s thirty to fifty percent too low. It isn’t dishonesty — memory simply isn’t built for adding up a hundred small amounts.

Here’s how to arrive at a real number, and what usually distorts it.

Why estimates fail

Three mechanisms, all pushing the same direction:

You remember big shops and forget small ones. A 350 zł Saturday shop sticks. Eight trips to the corner shop at 15–25 zł each don’t — and those often add up to more over a month.

You don’t count eating out as “food”. Lunch at work, coffee on the way in, Friday pizza all get filed mentally under “going out” rather than “food”. Budgetally they’re the same expense.

Groceries are mixed in with everything else. One supermarket receipt has tomatoes, laundry powder, a lightbulb and a toy on it. Counting the whole total as food overstates it; skipping the receipt understates it.

What to actually count

Before you start, decide what’s in the category — otherwise in three months you’ll be comparing two different things:

  • Groceries — food and soft drinks bought to eat at home.
  • Eating out — restaurants, delivery, work lunches, takeaway coffee.
  • Alcohol — worth separating, because it’s often large and very unevenly distributed.
  • Household and cleaning — not food, even though it comes from the same shop.

The most common mistake is dumping everything from the supermarket into one bucket. If the question is “what does food cost me”, the laundry powder has to come out.

Doing it without a spreadsheet

Transcribing receipts by hand works for two weeks and then stops. That’s not a discipline failure, it’s just what happens.

The realistic minimum is photographing every receipt. Do it in Paragonix and the lines are stored as data — name, price, shop, date — rather than as an image that’s unreadable in a month. That’s enough for anything to be calculated later.

We’re working on summaries that turn this into charts automatically; why that’s harder than it looks is covered in the post on Statistix.

Three numbers worth knowing

Once you have a month of data, work out:

  1. Cost per day. The total divided by days. Easier to compare than a monthly figure, and it immediately shows what a “normal” day costs.
  2. The share spent eating out. Usually the most surprising number. Three work lunches a week at 35 zł is over 400 zł a month — as much as a substantial shop.
  3. The number of shop visits. Not the amount, the count. Frequent small shops cost more than infrequent large ones, because something off-list lands in the basket every single time.

What to do about it

Simply knowing the number changes behaviour more than any budget does. But if you want specifics, two things give the largest effect for the least effort:

  • Cut the number of visits, not the size of the basket. One larger weekly shop instead of four small ones typically drops the total by low double-digit percentages with no sacrifice at all.
  • Check that discounts actually applied. A shelf promotion isn’t always in the system — where to look is covered in the guide to discounts.

If you live with someone and some shopping is shared, separating your part from theirs is its own problem — the flatmates guide covers it.

Splitting a receipt right now?

Snap it, share the link, and everyone claims what they ordered. Free, no account needed.

Scan a receipt now