All blogs

Reporting

How to read a variance report

Most teams glance at the variance total, wince, and close the tab. Read properly, the same page tells you exactly where the money is leaking, and which fix to make first.

FA

Folake Adeyemi

Customer success · · 5 min read

The variance report has a reputation problem. It arrives after a count, it's usually bad news, and it reads like an accusation. So it gets the same treatment everywhere: a glance at the total, a sigh, and a click somewhere less uncomfortable.

That's a waste of the most diagnostic page in the system. A variance report isn't a scorecard. It's a map of where your process leaks. You just have to read it in the right order.

Three numbers, one page

Start with three readings, in this order. The variance value tells you what the gap costs, and it decides how much attention the problem deserves. The variance rate, as a percentage of stock counted, tells you whether things are getting better or worse, and it's the only number worth trending week over week. And concentration tells you where to look: sort by absolute variance and see how much of the total sits in the top five SKUs.

That last one is the reading most teams skip, and it's the one that changes behaviour. Variance is almost never spread evenly. When four products carry eighty percent of the gap, you don't have a stock problem. You have four product problems, and those are fixable this week.

Reading the shape

Once you know where the variance sits, its shape tells you why it's there.

  • One SKU, repeatedly negative: a unit problem. You're selling it in a measure the system doesn't know about: pours, portions, halves.
  • A whole category off at once: a receiving problem. A delivery was booked in as ordered, not as delivered.
  • Everything slightly off: a timing problem. The count happened while the shop was trading.
  • Negative on the shelf, positive in the stockroom: a transfer problem. Stock moved, the paperwork didn't.

None of these are mysteries, and none of them are fixed by counting harder. Each shape points at one step in the process, and the fix lives there, not in the stockroom.

From reading to fixing

Pick the single largest contributor, name the shape, and fix that one cause before the next count. One fix per cycle is enough, because variance work compounds. Teams that do this watch the rate fall for six or seven consecutive counts, not because they got stricter, but because each report had one less thing to say.

Share this post

Stocked is the system these notes are written from. Start free or read more from the blog.

Read next

All posts