How to Calculate Pour Variance at Your Venue

Pour variance is one of the most useful numbers a bar manager can track, and — like retail conversion rate — one that's very difficult to measure accurately without dedicated tools. Getting a handle on pour variance turns "we think we're losing some stock somewhere" into a specific, trackable percentage a manager can actually act on, rather than a vague monthly write-off.

The pour variance formula

Pour variance is calculated as: ((expected consumption based on sales − actual consumption poured) ÷ expected consumption) × 100. If POS data says 100 standard measures of a spirit should have been poured based on what was rung up, but bottle-level data shows 112 measures actually left the bottle, that's a 12% negative variance — more product going out than being sold for. A positive variance (less poured than sold) usually points to a different problem: a data or till-recording issue rather than a stock-loss one.

Common causes of variance

The most common driver is simple over-pouring — free-hand pours running slightly heavier than the standard measure, which adds up fast across a busy shift. Comped or spilled drinks that never get rung through the till are another major contributor, along with product used in cocktails that draws down more than a single standard measure without being logged as such. A smaller but real factor is straightforward theft or unauthorised consumption, which pour-level tracking makes far more visible than a monthly stocktake ever could.

Variance looks different by product type

A high-volume, low-price spirit typically tolerates a slightly higher variance percentage before it's worth investigating, simply because the dollar value per pour is small. A premium spirit or a wine sold by the glass deserves scrutiny at a much lower variance threshold, since the same percentage gap represents a far larger dollar loss per bottle. Setting a single venue-wide variance threshold for every product ignores this — it either chases negligible variance on cheap spirits too aggressively, or lets meaningful premium-product loss slide under a threshold set for cheaper stock.

Why manual tracking rarely catches it in time

Calculating true pour variance by hand requires an accurate, pour-by-pour consumption record matched against sales data for the same period — something a monthly physical stocktake simply can't provide with any real precision, since it only shows the accumulated gap weeks after it happened. This is exactly what a liquor management system automates: each pour is measured and logged at the bottle via a wireless spout, matched continuously against POS sales data, so variance is visible within days rather than discovered a month later during a stocktake reconciliation.

What counts as a normal variance

Some variance is expected in any bar — a small amount of spillage, testing pours, and legitimate comps are a normal part of running a venue, not a sign of a problem. What matters is tracking your own venue's baseline variance consistently, so a genuine spike (a new staff member, a new product line, a specific shift) stands out clearly against what's normal for your bar, rather than being lost in month-to-month noise from an inconsistent measurement method.

Acting on the number

Once pour variance is tracked reliably by bar, waiter or venue, it becomes a targeted diagnostic tool rather than a vague monthly write-off — a spike traceable to one specific bar or shift is a concrete, investigable lead, in a way a single venue-wide stocktake number never is.