How do I compare sales across multiple bars fairly?
By Better Bartender · Published
Separate growth from new locations from changes at the same venues, then compare complete periods using consistent definitions. A group growing from £20,000 to £24,000 can still have declining like-for-like sales if the increase comes from a new bar.
Define the comparison cohort
A like-for-like comparison needs an explicit rule for which venues qualify in both periods. Start with locations that traded and supplied complete, comparable information in each window. Report excluded locations and their reason separately. Do not quietly remove poor performers or missing sites to make the headline look better.
A complete report still may not represent comparable trading. Note renovations, restricted hours, changed concepts and exceptional events. Your definition can differ from another operator’s, so publish the rule alongside the number.
A worked two-venue example
All figures below are invented and use one currency and one consistent sales basis.
- Venue A records £12,000 in the first period and £13,000 in the second.
- Venue B records £8,000 and then £7,000.
- A newly opened Venue C records £4,000 only in the second period.
The group total rises from £20,000 to £24,000: 20% growth. The continuing A-and-B cohort remains at £20,000: 0% like-for-like growth. Venue A improves while Venue B declines. All three statements matter; the group headline alone conceals the difference.
Why not average the percentages?
Venue A grows 8.33%; Venue B falls 12.5%. Their simple average is about −2.08%, but the combined sales change is zero. For a sales-weighted group result, compare summed comparable sales: (current cohort sales − prior cohort sales) ÷ prior cohort sales. An unweighted average answers a different question about the typical venue percentage.
When prior sales are zero, show the absolute change instead of an undefined growth percentage. If the earlier period is missing, label the comparison unavailable. Returns or corrections may require a separate explanation before a percentage makes commercial sense.
Normalise only when the inputs support it
Sales per open hour can add context when opening hours differ, but revenue is not uniform throughout a shift. Extra quiet hours and lost peak hours are not equivalent. Treat the ratio as another view, not a correction that makes all venues identical.
Likewise, transactions, recorded serves and customers are different denominators. A group with inconsistent POS buttons or measure sizes needs product mapping before item-level comparisons are trustworthy. Keep original labels and mappings traceable.
What belongs in a group review?
Show total sales, like-for-like sales, venue contributions to the change, reporting coverage and the operational notes that explain exclusions. Then use a weekly venue review for the sites that need attention. Rank a venue only after checking whether the comparison fits its service model.
Better Bartender brings AI-native reporting to the decisions behind a stronger hospitality business. Tell us about your venues.