Perpetual inventory keeps a running balance that updates with every transaction, which means it is only as honest as the data feeding it. Periodic inventory physically counts what exists at set intervals. A weekly count is periodic counting done tightly, and TrackItWeekly is built around that structure: count, review usage, decide the order.
Perpetual inventory starts from a baseline and adjusts it continuously. Every sale, every delivery, every transfer is supposed to flow in and change the balance. In theory you always know what you have, because the system never sleeps.
Periodic inventory does not pretend to know continuously. At a chosen interval, someone walks the shelves and counts what is physically there, and that count becomes the new truth. Daily, weekly, monthly, or once a year, the interval is the design decision.
| Dimension | Perpetual | Periodic |
|---|---|---|
| Best environment | High transaction volume, digital-only movements | Physical stock, waste, transfers, human hands |
| Trust model | Trust the data feeds | Trust the count |
| What it misses | Everything that never got recorded | Nothing on the shelf, but only on count day |
| Failure mode | Silent drift, then a surprise reconciliation | Stale picture, if the interval is too long |
| Typical cost and setup | Integration and data discipline required | A counting method and a cadence |
Because the register only sees the register. Waste, spoilage, unrecorded transfers, employee meals, breakage, and the case that never made it off the dock never enter the feed. The balance keeps adjusting itself confidently around a reality it cannot see. Over months, the system's number and the shelf's number walk apart, and the gap is exactly where the money went.
That is why even fully perpetual operations still do physical counts. The count is not a rejection of the system, it is the calibration.
If you sell online, if nearly every unit movement is a recorded transaction, and if stock is picked and shipped rather than portioned by hand, perpetual is the right architecture and worth the integration work. Omnichannel retail is a perpetual problem, and pretending a weekly count solves it would be dishonest.
Weekly is the interval where periodic counting gets interesting for on-site sellers. Annual counts tell you how bad last year was. Monthly counts tell you how bad last month was. Weekly counts tell you what happened while you can still do something about it, and the week is short enough that usage patterns stay visible.
TrackItWeekly is built entirely around that structure. The count is the input: what is on hand, entered on a phone, offline if needed. The value is the decision: usage reviewed against your own history, PAR guidance from your rolling three-week average, and an order that starts from data. It does not pretend to be perpetual. It makes periodic so tight you rarely wish it were.
Many operators should. POS sales data answers what sold. The weekly physical count answers what exists. Compare the two and the variance is the most useful number in the building: it points at waste, theft, process gaps, or transfers nobody logged, while the trail is still warm.
If the count-to-decision rhythm is your missing piece, let us guide you. Start your first weekly count.