RevPASH

A restaurant has limited seats and opening hours, and RevPASH asks how much revenue those available seat-hours produced. It combines demand, duration and spending in one measure.

A seat-hour is one seat available for one hour, so 50 seats open for four hours provide 200 seat-hours, which makes capacity comparisons easier than raw revenue alone. For a simple whole-period calculation, divide defined restaurant revenue by seats available multiplied by hours open.

A fictional restaurant with 50 seats, four hours and $4,000 of sales has $20 per available seat-hour. The denominator counts available capacity, not only occupied chairs, so empty seats reduce the result, and a table that stays occupied longer also limits when new guests can sit, even if the first party spends well.

Cornell hospitality research describes RevPASH as a time-and-capacity measure and warns that assigning an entire check to the hour it opened can misrepresent short intervals when diners occupy seats across boundaries. Its study found large errors for short intervals in one restaurant, though that error size should not be generalised to every venue.

For an entire lunch period a revenue-over-total-seat-hours formula is perfectly useful, but for 15-minute or hourly comparisons you must decide how revenue is attributed across the time a meal occupies, and you should not call the first hour weak merely because checks were opened or closed later. Compare like periods, by day part or weekday, with the same seat count and sales treatment, because weather or a special menu can move demand without any change in seating policy.

Use actual available capacity for each period rather than a permanently posted maximum, since a private event may close a section or tables may be rearranged. If takeout revenue is included while takeout customers use no seats, RevPASH is inflated, so exclude it or label the figure as a blended measure, and record layout or opening-hour changes so that a before-and-after renovation comparison uses the updated denominator.

Revenue is not profit, and a high RevPASH may come with steep discounts, overtime or costly ingredients, so pair it with margin, labour and guest experience before deciding an offer worked. Average spend and table use both influence the metric: a short meal with modest spending can free a table for another party, while a longer meal may carry higher spending.

During a quiet afternoon more guests may raise RevPASH without displacing anyone, but during a peak a discount could attract early guests who occupy tables needed for later high-demand bookings, so model capacity before launching it. A reservation system can reveal dining duration and point-of-sale records show check timing and sales, though neither is perfect if arrival, seating, ordering and payment timestamps differ.

A manager can use a heat map of intervals to spot patterns, but a low number alone does not say whether to change price, staffing or opening hours. For a fictional cafe, a longer weekend brunch may have higher sales but lower RevPASH than a short, busy lunch, and the cafe might accept that because brunch attracts loyal customers or produces a better margin.