There are only two ways to know what’s on your shelves. You can count it on a schedule, or you can track every movement as it happens. Everything else is a variation on those two ideas.
The first is a periodic system. The second is perpetual. The choice between them decides how much software you buy, how much staff time counting eats, how accurate your numbers are on a Wednesday afternoon, and how quickly you find out that a product is about to run out.
Most comparisons of the two read like a sales pitch for software. This one won’t, because for a lot of operations the honest answer is that the cheaper method is still the right one. What follows is the decision laid out on the terms that actually matter: order volume, SKU count, staff hours and cost.
Key Takeaways
- A periodic system updates inventory only at the end of a set period through a physical count. A perpetual system updates it continuously, with every sale and receipt.
- Periodic is cheaper and simpler but leaves you blind between counts. Perpetual costs more and demands scanning discipline, but answers “what do we have right now.”
- Spreadsheets still dominate: 84.8% of operators surveyed by inFlow in 2026 track inventory in them, and 74.3% use them as their only or primary tool.
- The real decision drivers are SKU count, how perishable your stock is, and whether you sell through more than one channel, not company size.
- Hybrid setups are normal and often correct: perpetual tracking on your top items, periodic counting on the long tail.
Save 80% of delivery management time
We handle everything:
- Dedicated operations manager
- Real-time tracking dashboard
- Automated customer notifications
- Urgent issue resolution
How each method works
The two systems differ in when your inventory record changes, and everything else follows from that.
A periodic system records purchases as they arrive and sales as they happen, but never adjusts the inventory quantity in between. On count day you physically count what’s left, value it, and work out what you must have sold by subtraction: beginning inventory plus purchases minus ending inventory equals cost of goods sold. If you want the full mechanics of the count cycle and that calculation, we cover it step by step in our guide to how the periodic inventory system works.
A perpetual system adjusts the record at the moment of each transaction. A case is scanned in at receiving and the count goes up. An order ships and the count goes down. Cost of goods sold accumulates continuously rather than being derived at the end. You still count physically, but the count’s job changes: instead of producing the number, it verifies the number the system already has.
That shift in the count’s purpose is the clearest way to tell the two apart. In a periodic system, the physical count is your data. In a perpetual system, it’s an audit of your data.
Periodic vs perpetual inventory at a glance
| Periodic | Perpetual | |
|---|---|---|
| Record updates | At period end only | Continuously, per transaction |
| Source of the number | The physical count | The transaction log |
| Stock position mid-period | Estimated from memory | Available on demand |
| Cost of goods sold | Calculated at close | Accumulates in real time |
| Setup cost | Minimal — a spreadsheet works | Software, scanners, barcoding, setup |
| Ongoing staff time | Concentrated in count events | Spread across every transaction |
| Scanning discipline required | None | High, and it’s unforgiving |
| Shrinkage visibility | Hidden inside COGS | Shows up as a variance |
| Interim financials | Incomplete by design | Usable at any point |
| Breaks down when | SKUs, locations or channels multiply | Staff skip scans |
The last row of that table is the one people underestimate in both directions. A periodic system degrades gradually as complexity grows. A perpetual system degrades suddenly, the moment the scanning stops being reliable. A perpetual system with bad discipline is worse than a periodic one, because it reports a precise number that happens to be wrong, and people trust it.
Do most companies use periodic or perpetual inventory?
Large retailers run perpetual systems almost universally. Small and mid-sized operations mostly don’t, whatever the software marketing implies.
In a March 2026 survey of 400 inventory, warehouse, supply chain and operations professionals, 84.8% reported tracking inventory in spreadsheets, and 74.3% used spreadsheets as their sole or primary tool with no dedicated software underneath. Dedicated inventory software was in use at 22%, ERP systems at 19%, and 27% still kept paper records (inFlow Inventory, State of Inventory Management 2026, published September 11, 2026).
A spreadsheet can technically hold a perpetual record, but in practice most spreadsheet-run operations are periodic: the sheet gets reconciled against a count rather than updated on every transaction. So the market majority is still counting on a schedule.
One finding in the same survey deserves a second look. 92% of operators said they were satisfied with their current approach, and then named inventory accuracy as their number one improvement priority. People are content with the method and unhappy with the results, which usually means the method is being blamed for too little and the discipline around it for too much.
What each method costs in money and hours
Cost comparisons usually stop at the software licence, which is the smallest number in the equation.
A periodic system’s cost is labour, paid in lumps. A full count of a few hundred SKUs is an evening’s work for two people, after hours, at overtime rates, every period. The cost is predictable and it scales badly. Double your SKUs and you roughly double the count.
A perpetual system’s cost is setup plus discipline. Software subscription, scanners, barcode labelling every item, and the bigger expense: training people to scan consistently and building the habit. The counting labour drops but never hits zero, because you still cycle count to verify.
The crossover is less about revenue than about how many times a day your stock position changes. An operation shipping fifteen wholesale orders a week can hold its position in its head between counts. One shipping two hundred retail orders a week cannot, and the periodic method stops being cheap. It just moves the cost from counting hours into stockouts and dead stock.
Which brings up the cost nobody budgets for. IHL Group’s 2025 analysis put the global cost of inventory distortion, out-of-stocks and overstocks combined, at $1.73 trillion annually, or 6.5% of global retail sales, with $1.157 trillion of it attributable to out-of-stocks (IHL Group, September 2025). A method that saves you $400 a month in software and costs you two lost wholesale accounts a year is not the cheaper method.
Which is better, periodic or perpetual, for your operation?
Neither is better in the abstract. Run your own situation through these five questions and the answer usually becomes obvious.
1. How many SKUs do you carry? Under roughly 200 items that one person can count in an evening, periodic holds up fine. Past 500, counts get long enough that they slip, and slipped counts mean stale data. In the thousands, perpetual is the only workable answer.
2. How perishable is your stock? Produce, dairy, flowers and prepared food change value and condition daily. A method that reports monthly cannot manage a product with a six-day life. Perishability pushes hard toward continuous tracking regardless of how few SKUs you have.
3. How many channels and locations? One room, one sales channel: periodic works. The moment the same inventory is sold through a storefront, a wholesale order form and a website, you need a single live number or you will sell the same case twice.
4. What does a stockout actually cost you? If it costs one walk-in sale, periodic’s blind spot is survivable. If it costs a standing weekly wholesale account that took you a year to win, it isn’t. Judge the method by the worst thing it lets happen, not the average.
5. Will your team actually scan? This is the question that sinks perpetual implementations. If receiving happens at 5am by whoever is available, and nobody will reliably scan a case before it goes on a rack, a perpetual system will produce confident, wrong numbers. Honest answer required.
Three operations and the right call for each
A florist with 120 SKUs, one shop, mostly walk-in trade. Periodic, counting weekly. Perishability argues for perpetual, but the SKU count is small and the shop owner sees the cooler daily. Weekly counts plus a short list of daily checks on high-value stems covers it.
A wholesale bakery with 60 products, 40 standing accounts, daily delivery routes. Hybrid. Perpetual on the finished goods going out on routes, where double-selling a customer’s order is the expensive failure, and periodic monthly counts on ingredients and packaging.
A specialty food distributor with 2,400 SKUs across two warehouses. Perpetual, no real debate. Two locations and that many SKUs means a periodic count can’t be completed cleanly, and anything in transit on count night corrupts the result.
The hybrid approach most operations end up using
The periodic-or-perpetual framing is a bit of a false binary. In practice the right answer for a lot of businesses is both, split by how much each item matters.
The split comes from sorting your stock by value contribution. Typically around 20% of SKUs drive roughly 80% of revenue, and those items earn continuous tracking, frequent counts and close attention. The long tail of low-value items can wait for a scheduled count without costing you anything. ABC analysis is the method for making that split properly rather than by gut feel, and it’s the single most useful thing you can do if you’re staying periodic.
Whichever side you land on, the gap between counts needs covering. A periodic system is blind between counts by definition, and even a perpetual system doesn’t protect you from a supplier arriving two days late. The buffer that absorbs both is safety stock, sized from your actual demand variability and lead times. With a longer counting period you need more of it, and that extra inventory is a quantifiable cost of staying periodic.
One category that falls through the cracks of both systems: the crates, totes and insulated carriers that leave on a delivery route and are supposed to come back. Neither method tracks them unless you decide to, and they represent real capital sitting in customers’ back rooms. Our guide to reusable packaging covers how to keep count of what’s out on the route.
How to switch from periodic to perpetual without breaking things
If you’ve decided to move, the transition is where most of the risk sits.
- Start from a clean full count. A perpetual system inherits whatever error exists at launch and carries it forever. Count everything properly first.
- Barcode before you go live, not during. Half-labelled inventory produces exactly the inconsistent scanning that ruins perpetual records.
- Roll out by category, not all at once. Put your A items on perpetual tracking first. They’re the highest-value and the smallest number of SKUs, so you get most of the benefit for a fraction of the setup.
- Keep cycle counting permanently. Perpetual doesn’t eliminate counting, it changes what counting is for. Rotating partial counts are how you catch the drift between the record and the shelf.
- Watch the variances in month one. Early variances tell you where scanning discipline is failing while it’s still cheap to fix.
Frequently asked questions
What is an example of a perpetual inventory system?
A café using a point-of-sale system where every drink sold automatically deducts the milk, cups and beans in its recipe from inventory, and every delivery is scanned in at the back door. The owner can pull an accurate stock position at any moment without walking to the storeroom.
Can you use both periodic and perpetual inventory?
Yes, and plenty of operations do. The usual split is perpetual tracking on high-value or fast-moving items and periodic counting on everything else. Just be clear about which items sit in which group, and don’t let an item fall into neither.
Is a periodic inventory system acceptable for accounting?
Yes. It’s a recognised method producing valid financial statements, as long as the physical count is done properly and valued on a consistent basis. The limitation is managerial, not regulatory: your interim figures are estimates.
Does a perpetual system eliminate physical counting?
No. It changes the count’s purpose from producing your inventory number to verifying it. Operations running perpetual systems still cycle count continuously, because the record drifts from reality through miskeys, damage and theft no matter how good the software is.
Which method handles shrinkage better?
Perpetual, clearly. In a periodic system, anything lost to theft, spoilage or breakage is absorbed silently into cost of goods sold and is indistinguishable from product you sold. A perpetual system surfaces it as a variance between the expected and counted quantity, which is the first step to doing anything about it.
Making the call
If you carry a manageable number of SKUs in one location, sell through one channel, and have the cash better deployed somewhere other than inventory software, stay periodic and run it with discipline: count your valuable items more often than your cheap ones, log your spoilage, and hold a calculated buffer.
If your SKU count is climbing, your stock is perishable, or you’re selling the same inventory through more than one channel, the periodic method has already started costing you more than the software would. You just can’t see it on an invoice.
And if you’re somewhere in the middle, which most operations are, split the difference deliberately rather than by accident. Track the items that matter continuously, count the rest on a schedule, and spend the saved effort on the twenty percent of your stock list that’s paying for everything.