Daily stock count for restaurants: variance, waste and par levels

Food is cash sitting in a fridge. It spoils, gets over-portioned, goes out as a staff meal nobody wrote down, or arrives short from a supplier and is signed for anyway. Every one of those losses is invisible until you count, and the longer the gap between counts, the harder it is to work out what happened.

A daily count does not mean counting everything every day. It means counting a handful of items that matter most, quickly and consistently, and comparing what you find with what should be there.

Why count daily, and what to leave for the full count#

A full count of every shelf takes hours and is best done weekly or monthly, usually at period end, so you have a closing stock value for your accounts. It is thorough but slow to give feedback. If a problem started three weeks ago, a monthly count only shows the total damage.

A daily count of a short list takes minutes. It catches problems the day after they start, while staff still remember what happened, and it changes behaviour: people take more care with items they know will be counted tomorrow morning.

What to count: your "A items"#

Choose the items where a loss hurts most. Look at your purchases for the last month and pick out those that make up the largest share of spend, then add anything easy to take or easy to waste. For most kitchens that is a list of somewhere between ten and twenty items.

• High value: proteins such as beef, chicken, lamb, seafood; premium cheeses.
• High risk of loss: items that are portable, popular with staff, or easy to over-portion.
• Short shelf life: fresh fish, dairy, cut fruit, prepared sauces.
• Key sellers: the ingredients behind your best-selling dishes, where a small portioning drift multiplies quickly.

Review the list every so often. If a menu change makes an item unimportant, drop it; if a new dish relies on an expensive ingredient, add it.

How to count so the numbers mean something#

A count is only useful if it is taken the same way every time. Small inconsistencies create variances that are really just counting noise.

• Same time every day: before service starts or after it closes, never in the middle of prep.
• Blind counts: the person counting writes down what they find without seeing the expected figure, so they are not tempted to "find" the right answer.
• By storage location, in shelf order: walk-in, freezer, dry store, bar, line fridges. Count in the order things sit, so nothing is missed or counted twice.
• Units that match your recipes: if recipes use grams, count by weight; if they use pieces, count pieces. Converting at the end invites mistakes.
• Opened packs and prepped items count too: half a tray of marinated chicken in the line fridge is stock.

The formula: expected stock and variance#

"Sold through recipes" is the important part. If the till recorded 136 chicken plates and each uses 180 g of chicken, that is 24.48 kg of chicken that should have left the kitchen. This only works if your recipe cards are accurate, which is one reason recipe costing and stock counting belong together.

A negative variance means less stock than expected; a positive one means more. Both deserve a look. A positive variance often points to a recipe that overstates the quantity, or a delivery that was booked in twice.

A worked example

Here is one morning count for four items. All quantities and costs are illustrative, chosen only to show the method.

The total looks small, but it hides two separate stories. Chicken is down 1.2 kg, worth checking against portioning on the line. Salmon is up, which suggests the recipe quantity is higher than what the kitchen actually serves, or a receiving error. Netting them off would hide both. Look at each line, and value each one at cost rather than menu price.

Log waste, staff meals and comps before they become variance#

Anything that leaves stock without a record turns up as unexplained variance. The fix is to log it when it happens, with a reason: spoiled, dropped, overcooked, returned by a guest, past its use-by date. A waste sheet on the wall or a quick entry on the till both work, as long as it is done at the time and not reconstructed later.

Staff meals and comped dishes are not waste, but they are usage. Ring them through the till under their own category so the recipe deduction happens and the cost is visible. Separating these reasons matters: a rising waste figure for "spoiled" points to ordering, while a rising figure for "overcooked" points to training.

Common causes of variance#

1. 1Unrecorded waste: the most common cause, and the easiest to fix.
2. 2Portioning: generous scoops and heavy hands, especially when it is busy. Scales and portion tools at the station help.
3. 3Wrong or outdated recipe: the dish changed, the card did not.
4. 4Receiving errors: short deliveries signed as complete, or the wrong pack size booked in. Check weights and counts at the door.
5. 5Theft: possible, but treat it as the last explanation, not the first. Rule out the process causes, look for patterns over several days, and handle any concern privately and fairly. An accusation built on a single bad count damages trust and is often wrong.

Par levels: turning counts into orders#

The same count that measures variance also tells you what to order. A par level is the quantity you want on hand after a delivery so you do not run out before the next one.

Using the illustrative chicken figures: if the kitchen uses about 25 kg a day, delivery comes every two days, and you keep 8 kg of safety stock, the par is 25 × 2 + 8 = 58 kg. With 11.3 kg counted and nothing on order, you need 46.7 kg, which becomes five 10 kg cases. Revisit pars when sales patterns shift, for example around Ramadan, school holidays or summer.

EXO NODES records counts, waste and variance against recipe usage from the till, and books the variance to the ledger, so the stock figure in your accounts matches what is on the shelves.

A simple daily routine#

1. 1Count the A-item list at the same time, by location, blind.
2. 2Check that yesterday’s deliveries and waste entries are all recorded.
3. 3Compare counted with expected and value each variance at cost.
4. 4Investigate any line that is out of the ordinary, while the shift still remembers.
5. 5Set today’s orders from par levels.
6. 6Note what you found and what you changed, so patterns show up over the week.

Key takeaways#

• Count a short list of high-value, high-risk items daily; keep full counts for weekly or month end.
• Count consistently: same time, blind, by location, in recipe units.
• Expected = opening + received − sold via recipes − recorded waste; variance = counted − expected, valued at cost.
• Log waste, staff meals and comps with reasons at the time they happen.
• Rule out process causes before suspecting theft, and handle concerns fairly.
• Par = daily usage × days to next delivery + safety stock; order up to par in pack sizes.