Par Levels Explained: How to Set Reorder Points That Actually Work
Parly Team路February 18, 2026路5 min read
What is a par level?
Stock level zones bar
A par level is the minimum quantity of an item you should have on hand at any given time. When your stock drops to or below this number, it is time to reorder.
Think of it as a trigger. Not "how much should I order?" but "when should I start ordering?" The par level answers the second question. Order quantities come after.
Why most cafes get par levels wrong
The most common approach to setting par levels is asking the person who does the ordering: "How much do we usually keep around?" Their answer becomes the number. The problem is that this number is based on memory, habit, and anxiety, not data.
Results of bad par levels:
- Too high: You over-order, tie up cash in excess inventory, increase waste from expiration, and run out of storage space.
- Too low: You stockout during busy periods, make emergency runs, lose sales, and frustrate customers.
- Static: A par level set in January might be wrong by March. Seasonal shifts, menu changes, and volume trends all change consumption patterns.
The formula
Par level formula visualization
A solid par level accounts for three things: how much you use, how long it takes to get more, and a buffer for variability.
Par Level = (Daily Usage x Lead Time) + Safety Stock
馃挕 Better inputs, better par levels
Daily usage calculated from POS sales data and recipes is more accurate than estimates from memory. If you have a POS connection, use recipe-based consumption rates as your starting point and validate against physical counts.
Let's break that down.
Daily usage is your average daily consumption of an item. Calculate this from actual count data, not estimates. Take your total usage over a period (say, 14 days) and divide by the number of days.
Example: You used 42 cartons of oat milk over 14 days. Daily usage = 3 cartons/day.
Lead time is the number of days between placing an order and receiving it. This varies by supplier. Your dairy distributor might deliver next day. Your specialty matcha supplier might take 5-7 business days.
Example: Oat milk lead time = 1 day (next-day delivery from Metro Supply Co).
Safety stock is your buffer against variability. Weekends are busier than Tuesdays. Sometimes a delivery is late. Sometimes you have an unexpectedly busy day. Safety stock absorbs these shocks.
A simple safety stock formula: take your highest single-day usage in the past two weeks and subtract your average daily usage. Multiply by lead time.
Example: Your busiest day used 5 cartons of oat milk. Average is 3. Safety stock = (5 - 3) x 1 = 2 cartons.
Par level for oat milk = (3 x 1) + 2 = 5 cartons.
When your count shows 5 or fewer cartons, it is time to order.
Adjusting for delivery schedules
Delivery timeline with stock curve
Not every supplier delivers every day. If your coffee roaster only delivers on Wednesdays, your lead time is not "2 days." It is "the number of days until the next possible delivery."
This means your par level for coffee beans needs to cover the maximum gap between deliveries. If you order on Monday for Wednesday delivery, but your next order after that arrives the following Wednesday, you need 7 days of supply, not 2.
For items with infrequent delivery windows:
Par Level = (Daily Usage x Days Between Deliveries) + Safety Stock
This is where many cafes under-order. They calculate based on lead time (2 days) when they should calculate based on delivery frequency (7 days).
Category-specific guidance
Par level settings screen
Different item types need different approaches.
Perishables (milk, cream, pastries)
Short shelf life means par levels should be tight. Over-ordering means waste. Under-ordering means stockouts. Count these daily and keep par levels at 1-2 days of supply above your daily usage.
For pastries with standing orders, your par level is effectively set by your daily order quantity. The adjustment happens in the standing order itself, not in a reorder trigger.
Dry goods and paper products (cups, lids, sugar, tea)
These have long shelf life so the cost of over-ordering is lower (storage space and cash flow, not waste). Set par levels with a more generous safety stock. Running out of cups during Saturday morning rush is a much worse outcome than having an extra case in the back.
Specialty and long-lead items (matcha, specialty syrups, chai)
These are the highest-risk items because lead times are long (5-7 business days) and you cannot get emergency replacements. Par levels should include 50-100% safety stock above the formula result.
If your matcha takes 7 days to arrive and you use 2 tins per week, your base par is 2 tins. But add at least 1-2 extra tins as safety stock. Running out of your signature drink for a week is not an option.
When to recalculate
Par levels are not permanent. Review and adjust them:
- Monthly for stable items with consistent usage
- After menu changes that affect an ingredient's demand
- At seasonal transitions (iced drink season vs. hot drink season shifts consumption dramatically)
- After any stockout to determine if the par level was too low
- After significant waste to determine if the par level was too high
From par levels to automated ordering
Once you have reliable par levels, the ordering decision becomes mechanical: current stock minus par level equals surplus (or deficit). If you are at or below par, order. If you are above par, skip.
The order quantity itself is a separate calculation: how much to bring stock up to your target level (usually par level plus a few days of supply). But the trigger, the "should I order?" question, is answered by the par level alone.
Good par levels remove the biggest source of ordering anxiety. You stop wondering "did I order enough?" and "should I add a few extra just in case?" The data answers those questions for you.
Start with your top 10 highest-volume items. Calculate their par levels using two weeks of count data. Use those for a month, track the results, and adjust. Then expand to the rest of your inventory. Within two months, every order you place will be backed by data instead of intuition.