Revenue Prediction and Menu Analysis

Forecasting Process

• Record actual counts and other data to evaluate forecast accuracy and adjust future forecasting.
• Adjust prior data using rolling averages or trends.
• Collect historical data.
• Make adjustments based on qualitative factors to create a final forecast.

Data Collection

• Historical Quantitative Data:
  • POS system
  • Reservation/walk-in book
• Qualitative Data:
  • Record daily in a log or POS system.
  • Examples:
    • Weather
    • Special events
    • Conventions
    • Holidays
    • Construction projects
    • Competition changes
    • Marketing promotions
    • Anything that might explain business fluctuations

Trends

• Start with the customer count from the same day of the week from the prior week or the prior year.
• Adjust based on historical trend percent increase or decrease.
• Formula:
  % \text{ change} = \frac{\text{Current customer count} - \text{Prior count}}{\text{Prior count}}

Example

• Restaurant saw 347 customers the first Friday in March this year and 360 the first Friday in March last year.
  • Calculate the percent change from last year:
    % \text{ change} = \frac{347 - 360}{360} = -0.036 \text{ (3.6\% decrease)}

Using Trends

• Trends are only significant if repeated over and over each same day of the week.
• Manager uses the trend to calculate an initial forecast:
  \text{New Customer Count} = \text{Prior Count} \times (1 + \text{percent change in decimal form})

Example

• Business has a consistent trend of 1.4% increase over last year’s customer counts.
• If the first Friday in March last year saw 340 customers, how many should be forecast for the first Friday in March this year?
  \text{New Count} = 340 \times (1 + 0.014) = 344.76 = 345

Rolling Averages

• Use rolling averages when no trend is consistent across weeks.
• Formula: \text{Rolling Average Count} = \frac{C1 + C2 + C3 + \ldots + CN}{N}
  • Where:
    • C = Customer count for that period
    • N = total number of periods counted

Example

• Business forecasts using rolling averages over 4 weeks. Guest counts for the past 4 weeks have been 418, 437, 398, and 414.
  • What is the forecast for the upcoming week?
    \text{Rolling Average} = \frac{418 + 437 + 398 + 414}{4} = 416.75 = 417

Adjusting for Qualitative Data

• Adjust initial forecast based on qualitative data.
• Keep adjusting as new data comes in (weather report changes, for example).
• Forecast:
  • 1-2 weeks out for employee scheduling.
  • 2-3 days out for (most) ordering.
  • 1 day out for kitchen production schedule.

Evaluating the System

• Constantly compare the forecast to the actual guest count to learn from mistakes and to make more accurate forecasts in the future.

Forecasting Sales

• Converting customers to revenue helps to budget and control costs.
• Average Check:
  • Amount of revenue the average person generates on a check.
  • Calculate average check by server to identify strong servers.
  • Calculate average check by day, week, or meal period to inform when to offer promotions.

Average Check Formula:

\text{Average Check} = \frac{\text{Revenue for a Period}}{\text{Guests for that Period}}

Example

• Over the past 4 Mondays, a restaurant has served a total of 1,104 guests for lunch and brought in $12,385.00 from those 4 periods.
  • What is the average guest check for Monday lunch during this period?
    \text{Average Check} = \frac{$12,385.00}{1,104} = $11.218297 = $11.22

Uses for Average Check

• It is a control tool.
  • If it changes much over time, the manager should research why and correct problems or reinforce results.
• When seating is limited, increasing the average check may be the only way to improve profit.
• \text{Forecast Revenue} = \text{Forecast Guests} \times \text{Average Check}

Example

• Restaurant with an average check of $47.58 forecasts 3,700 guests next month.
  • How much revenue should the manager expect next month?
    \text{Forecast revenue} = 3,700 \times $47.58 = $176,046.00

Using Forecast Revenue

• Using forecast revenue and target food cost %, beverage cost %, and labor cost %, a manager can determine the budget in dollars for a given period of time.

Seat Turnover

• When seating is maxed out, a manager can serve more customers in a period by increasing the seat turnover or number of customers per seat in a given period.
• \text{Seat Turnover} = \frac{\text{Customers Served in a Period}}{\text{Total Seats in Dining Room}}

Menu Mix

• Menu Mix % = % of sales that come from each menu item.
  • It is usually divided by menu category.
  • It is relatively consistent for a meal period across the same days of the week.
    • Caveat: may change seasonally or with weather and definitely with menu change.

Menu Mix – Step 1

• Menu mix calculated against total within a menu category.
• Manager must calculate % of guests who purchase food from each menu category.
  % \text{ buying category} = \frac{\text{Guests buying that category}}{\text{Total guests}}

Example

• Of the 330 guests in a restaurant one night, only 140 buy dessert.
  • What percent bought dessert?
    % \text{ buying dessert} = \frac{140}{330} = 0.42424242 = 42.42\%

Menu Mix Formula

• Note: Number of each item sold and total sold come from POS system or sales receipts
  \text{Menu Mix %}= \frac{\text{Number of that item sold}}{\text{Total number of items sold}}

Example

• Restaurant typically sells 140 desserts each Monday. Of those, 37 are usually sorbet.
  • What percent of desserts is the sorbet?
    \text{Menu Mix %} = \frac{37}{140} = 0.2643 = 26.43\%

Forecasting Number of Menu Items Sold from Menu Mix Percents

Step 1

Guest Forecast * Percent buying a category of food = Number of dishes in that category forecast to be sold

Step 2

Number of dishes in a category forecast to be sold (step 1) * Menu Mix % for an item = Number of that item forecast to be sold

Example

1. No. ordering dessert = 370 * 0.42 = 155.40
2. No. of sorbets = 155.4 * 0.37 = 57.498 = 58 sorbets

• Historically, 42% of guests order dessert Thursday night. Of those, 37% are sorbet.
  • How many sorbets should the pastry chef plan for next Thursday if the guest forecast is 370 guests?

Forecasting Kitchen Production

• Manager uses forecast menu mix sales to plan kitchen production.
• Manager should adjust forecast based on qualitative data, desire to have a buffer, or desire to run out of certain foods to avoid leftovers.
• Accurate production schedule minimizes leftovers, waste, purchases, and labor costs.

Menu Analysis for Increased Profitability

• Making the most profitable items the most popular ones helps to maximize overall profit.
• Menu Analysis is best done over a long time period (several months or a year).
• Menu Analysis is a process through which managers compare each menu item’s profitability and popularity.

Calculating Popularity

• Work with only one menu category at a time…
  1. Start with the number of each item sold (in one category); add total items sold and divide by the number of menu items in that category. This is “average number of each menu item sold.”
  2. Popularity Benchmark = average # of each menu item sold * 70% or 0.7
  3. High popularity items sell more items than the benchmark; low popularity items sell fewer items than the benchmark.

Calculating Profitability

• Work with only one menu category at a time…
• Contribution Margin (CM) = Item Sales Price – Item Cost of Goods
• Menu CM = # sold (for an item) * CM (for that item)
• For each item, calculate item CM = item sales price – item food cost.
• For each item, calculate Menu CM = number of items sold * its item CM. Add all menu CM’s together to get the menu CM total.
• Calculate Average Weighted Menu CM = total menu CM ÷ total number of items sold (in that menu category)
• An item is “high profitability” if its item CM is higher than the average weighted menu CM. An item is “low profitability” if its item CM is lower than the average weighted menu CM.

Menu Analysis Categories

• Star: High popularity, high profitability. Leave these items alone.
• Plowhorse: High popularity, low profitability. May increase sales price, reduce food cost, or leave alone if it is a signature dish/draw.
• Puzzle: Low popularity, high profitability. Relocate on the menu, highlight on the menu, rework menu description, suggestive sell.
• Dog: Low popularity, low profitability. Often requires a major change: increase sales price, suggestive sell, or replace it with a dish that still meets the needs of the customers who used to order the dog (e.g., replace a vegetarian dish with another vegetarian dish).

Reconciling Kitchen Production with Sales

• Kitchen production schedules can be used to reconcile food produced against food sold.
• All food prepared by the kitchen must be accounted for to protect against theft.
• Kitchen Production Sheet uses menu mix % and forecast to predict how many of each dish will be sold.

Example

• Restaurant forecasts 130 guests for dinner tomorrow. 42% of guests usually order dessert.
  • Guests ordering dessert = 130 * 0.42 = 54.60 = 55 guests ordering dessert

Example Chart