Coffee Shop Inventory: The Owner's Full Guide

Coffee shop inventory has a reputation problem. It sounds like a chore you do because a bookkeeper told you to, filed next to filing.

It is actually the input to every operating decision you make. What to order Tuesday, whether the new drink is worth its shelf space, why food cost climbed two points last month: all of it comes out of a count. A cafe that counts well can answer those in minutes. A cafe that does not is running on memory and finding out at the worst possible time, usually at 9:40 on a Saturday with a line at the register.

This is the full loop, in the order you would actually build it.

What counts as inventory in a cafe

Start by cutting the list down. A cafe is not a restaurant, and a 400-line item list copied from a restaurant template dies in week two.

Three groups carry almost all of the money and nearly all of the risk:

• Beverage. Beans, matcha, chai, milks and alternative milks, syrups, teas, anything that goes in a cup. This is where the volume and the spoilage live.
• Paper and plastics. Cups, lids, sleeves, straws, napkins, filters, bags. Cheap per unit, brutal when you run out, and the category most likely to be ordered by vibes.
• Chemicals and supplies. Sanitizer, soap, gloves, towels, trash bags. Low glamour, high embarrassment when missing.

Pastries usually sit outside the count because they arrive and sell within the same day or two; they get tracked as deliveries and waste instead of counted on a shelf.

Sixty or so items covers most single-location cafes. If your list is twice that, you are probably counting things whose absence would never change a decision, and every one of those items is a small tax on every count you will ever run.

How to actually take the count

The count is a walk, not a spreadsheet exercise.

Walk the stockroom in physical order and record what you see. Do not alphabetize, do not group by supplier, do not bounce between shelves because that is how the list was typed. Shelf order is the single biggest speed difference between a count that takes 10 minutes and one that takes 40. A recent full count at my shop was 58 items in under 10 minutes on a phone, and the phone matters: a clipboard means transcribing later, and transcription is where numbers go wrong.

The rest of the method:

• Count the same units you buy in. If you order oat milk by the case of twelve, decide once whether you are counting cases or bottles and never mix them mid-count.
• Count at the same time of day. A Monday morning count and a Monday closing count are different numbers, and comparing them tells you nothing.
• Write down zero. An empty shelf is data. Skipping it makes it look like nobody checked.
• Do not fix problems mid-count. Note the weird number, keep walking, investigate after. Counts that turn into investigations never finish.

The step-by-step version, with the phone workflow, lives in the practical counting guide, and if you want the paper version to start, take the count sheet template.

How often to count

Three times a week fits a cafe well: Monday, Wednesday, Friday.

Monday reads the weekend, which is where the variance is. Wednesday catches drift early enough to fix the rest of the week. Friday sets up the weekend, especially with suppliers who do not deliver Sunday. That cadence gives you enough points to see a trend without turning counting into a second job.

A few items earn daily attention regardless: milk, anything perishable with a one-day life, and whatever your best-selling drink depends on. Everything else is fine on the three-day rhythm. If you are choosing a starting point, start with Monday and add days once the routine sticks, because handing the count off works better when the habit is small and boring than when it is ambitious and skipped.

Turning counts into orders

A count by itself is a number on a screen. The order is where it pays.

The arithmetic is simple: usage per day, times days until the next delivery, minus what is on the shelf, plus a buffer. The buffer is your par level, and writing it down once beats re-deciding it every week under time pressure.

Then fit it to the suppliers, because their calendars, not yours, decide when the decision is due. Every supplier has an order deadline and a delivery rhythm, and ordering against those windows is what separates a calm Tuesday from an emergency. When the count is current, the order writes itself from the count and your job is editing it, not building it.

Both mistakes cost money in opposite directions. Over-ordering parks cash on a shelf until some of it expires. Under-ordering sends you out at retail prices mid-shift, which is how a $40 milk run replaces a $12 case. Long-lead items are the least forgiving; matcha needs to be ordered on its lead time, not on the day you notice the tin is light.

When your counts and your sales disagree

Eventually you will count something and the number will be wrong in a way you cannot explain. This is normal, and it is information.

You sold 180 lattes, which should have used about 12 gallons of milk, and 15 are missing. The gap is real and it has causes: overpouring, spills, staff drinks, comps, and most commonly, drinks that used a different milk than the recipe assumes. Modifiers are the usual culprit, and the mismatch between counted and expected is worth chasing precisely because the answer is usually a fixable habit rather than a mystery.

The gap is also where waste hides. Nothing rings when milk goes down the drain. Subtract sales from usage over a few weeks and waste stops being a vague guilt and turns into a short list of items and dollars, which is a problem you can actually solve.

What inventory tells you about money

Once counts are steady, they stop being about stock and start being about margin.

Usage plus prices gives you what each drink costs to make, and that is the number under everything else: your real food cost percentage, prices set on purpose, and the frequent surprise that your best seller is not your best earner. When food cost drifts up, counts are how you find which item moved instead of blaming the whole menu.

Usage over time also predicts. Thirty days of counts and sales tell you what next week needs, and what next month looks like, which is the difference between ordering for the season and reacting to it.

Spreadsheet, Square, or a system

Most cafes start in a Google Sheet, and that is a fine place to start. The sheet stops working at a predictable point: when the counts are current but nobody has time to turn them into an order, or when two people edit it and neither trusts the result. That failure mode is worth recognizing early.

Square is already tracking what sold, which is half the picture and the half most owners assume is the whole thing. It tracks items, not ingredients, so it can tell you 40 lattes sold and not that you have two gallons of milk left. Square's recipe features close some of that gap, with real limits worth knowing before you build on them. Whether you need anything beyond Square depends on how much the ingredient layer costs you today, and that is an honest question with a real "sometimes no" answer.

Wherever you land, the sequence does not change: count in shelf order, keep pars written down, order against supplier deadlines, and read the gap between sold and used. Start Monday with a phone and sixty items. The rest is built on that walk.