|
|
For this DEMO we return to "The Nighthawks Cafe" The table below (Table 1) shows some historical operational information for "The Nighthawks Cafe", which is a small up scale restaurant which now serves both lunch dinner and does occasional catering-banquets for nearby office buildings.
Although a not a large restaurant operation, "The Nighthawks" has been quite successful but both the chef and the general manager feel that the operation would be more efficient, profitable and better prepared to consistently offer their high level of customer service if they could better forecast their business.
"The Nighthawks Cafe" has, over the years, kept a good history of its covers (sales), but has not put those numbers to work in an organized, systematic and tactical manner.
That sales/covers history numerical data is what this DEMO is based on. You will find that data in the various tables below.
GENERAL BACKGROUND OPERATING INFORMATION
| Type of Operation | Casual Service · lunch, dinner, carryout, catering |
| Location | Chicago, IL & · near a business district and a shopping mall |
| Lunch | from 11:30 am to 2:30 pm Monday - Sunday |
| Dinner | from 6:00 pm to 11:00 pm Monday - Sunday |
| Days Closed | Christmass Eve (dinner) || Christmass Day (lunch and dinner) || New year Day (lunch and dinner) |
| Seating Capacity | 85 | |
| p |
| Dinning Room Turns | Maximum = 2 // Average = 1.4 // Acceptable = 1.35 |
| Sales/Covers History | Using up to the Last 5 years |
| Busiest to Slowest Day of the Week | Friday · Saturday · Thursday · Wednesday · Sunday · Monday · Tuesday |
| Busiest to Slowest Month of the Year | May · June · July · April · August · October · September · November · December · March · January · February |
DAILY COVERS: LAST 1 YEAR (364 samples)
| High | Low | Average |
|---|
| LUNCH | 184 | 67 | 120 |
| DINNER | 212 | 62 | 118 |
WEEKLY COVERS: LAST 2 YEARS (104 samples)
| High | Low | Average |
|---|
| LUNCH | 1031 | 595 | 790 |
| DINNER | 1180 | 577 | 755 |
MONTHLY COVERS: LAST 5 YEARS (60 samples)
| High | Low | Average |
|---|
| LUNCH | 3371 | 1817 | 2583 |
| DINNER | 3910 | 1726 | 2779 |
AVERAGE HISTORIC DAILY COVERS - BASED ON THE ABOVE DATA
| DAY | # Covers LUNCH | # Covers DINNER | # Covers TOTAL |
|---|
| Monday | 133 | 109 | 242 |
| Tuesday | 133 | 109 | 242 |
| Wednesday | 133 | 109 | 242 |
| Thursday | 133 | 109 | 242 |
| Friday | 133 | 153 | 286 |
| Saturday | 90 | 153 | 243 |
| Sunday | 84 | 84 | 168 |
-------------------------------------------------------------------------------------------------------------------------------------------
A LITTLE BACKGROUND ABOUT RESTAURANT COVERS FORECASTING ::
The GOAL is to improve the quality of the restaurant's production decisions in a cost effective manner. By that I mean the method you choose to measure and forecast should allow you to compare the benefits from improved accuracy with the costs for obtaining that improvement.
A Covers forecast primary PURPOSE is directed towards improving Labor Scheduling and its secondary purpose is directed toward improving purchasing. Specifically, to better match labor supply to with sales demand. The method employeed in this DEMO is what I call "Conditional Forecasting". This method is based on a selected set of "Casual Variables" combined with Historical Data that is "smoothed" by applying expert opinion to weight the importance of the selected causual variables.
|
The STANDARD for controlling production volume is to determine and produce (for any menu item) the number of portions that are likely to be sold on a given day. Knowing this number with a reasonable degree of accuracy is essential to foodservice entities so that intelligent plans can be made for purchasing and production. If you have a sales history of your menu items you know the percentage or ratio that each one sells relative to the others and the total number of covers. But before you can apply those individual percentages you need to have a accurate forecast of the total number of covers to apply those percentages/ratios against for production and preparation purposes.
As a harmless appearing example, let's say your average daily Item Costs of overproduction is a mere $25.00. At the end of a 360 day operational year you will have taken $9,000.00 out of your cash flow to cover those unnecessary purchases regardless if they went (written off) to feed the employees or not.
|
The PROCESS is sequential and it asks like a good detective::
- What is going on
- Why it is going on
- What it may lead to
and the KEY can be referred to as good guessing, or what I would like to call an Educated Guess.
From the above process a Sales Forecast is DEVELOPED by::
- gathering data
- then making an informed educated guesses about what may change and what may stay the same
- the follow up by testing those guesses against reality.
-------------------------------------------------------------------------------------------------------------------------------------------
IMPORTANT POINTS REGARDING FORECASTING IN GENERAL ::
- In some respects, forecasting what people will do in the future is predicting how people will act on their various intentions. Intentions are measures of an individual's plans, goals, or expectations about what they will do in the future. To be considered a useful benefit, a method must provide forecasts that are more accurate than a randon, uninformed guess and at a cost that does not exceed the benefits derived.
-
To approach making an educated guess, at the minimum you will need to know the most relevant causal relationships. Combining that with theory is helpful. For example, it is helpful to know that (in economic theory) the normal market reaction to a decrease in price (of your product or service) will tend to lead to a increase in demand for it, and vice versa, all other things remaining and being equal.
-
Forecasting is concerned with what the future WILL look like. Planning is concerned with what it SHOULD look like.
-
The terms "forecast", "prediction" and "projection" are typically used interchangeably. Many forms of forecasts are considered conditional forecasts. This means that if event 1 happens then result ABC is likely, but if event 2 happens instead, then you can expect that result XYZ will probably occur instead.
The following is a selected list of forecasting principles that were used to construct this forecasting program:
- keep the model simple
- use all the relevant data you can get
- use theory, not the data, as a guide to selecting causal variables
- adjust the data for important events that occurred in the past
- give more importance to the most recent data than earlier data
- consider seasonality, but by very careful when employing its impact
- be conservative when the situation is uncertain (which is most of the time)
- in forecasting, try to avoid using the "common sense" or "judgemental "approach
it tends to insert distorting biases like optimism and overconfidence
-------------------------------------------------------------------------------------------------------------------------------------------
THE "THE NIGHTHAWKS CAFE" DEMO ::
This is a DEMO version, with all of the results and recommendations fully functional. The program itself has been left fully operable.
That is, all of the logic and mathematical calculations are implemented as is the full capability to modify and update the generated forecasts.
|
The following summarizes the complete Forecasting process.
- You enter your expected number of lunch and dinner covers
- You also enter the current state of the variables that can have an impact.
- Based on previous sales history figures, the program applies mathematical factors
- The Results Screen outputs an UNMODIFIED FORECAST number of lunch and dinner and total covers.
- You are emailed a detailed report. Make sure you archive and retain this report.
- You should use this detailed report and its suggestions to ADJUST the UNMODIFIED FORECAST based on experience and knowledge.
- You can ADJUST (if you care to do so), the UNMODIFIED FORECAST by clicking on the adjust unmodified forecast link available on the bottom of all of the pages.
- Finally, once you have the ACTUAL sales figures for the Forecast Date, you should UPDATE the UNMODIFIED FORECAST. You can do this by clicking on the Update/Edit a Forecast link available on the bottom of all of the pages and selecting the corresponding date from the list.
- Make sure that you try and explain any causes for the difference between the ACTUAL and the UNMODIFIED FORECAST. Enter you explanations into the OBSERVATIONS box. For example, you may have been way off on the weather forecast input, or perhaps a large group showed up without having made a reservation..
If the above process is adhered to. the next time the sales history numbers are analyzed, benchmarked and then applied, there is no question that the accuracy of the results will improve
|
|
