What is a sales or customer traffic forecast?
It is important to get your sales forecast right, and key to providing consistent exceptional customer service. If your forecast is right you can match staff levels to the customer demand. This will also lead to converting on all revenue opportunities.
Sales forecasting is done by projecting your future sales and customer traffic based on historical data.
The ingredients for generating the right sales forecast are:
- Historical source data – typically sales and a measure of customer count (e.g. transactions, items, orders)
- Understanding your growth index
- Know why and how to use a rolling average
- Consider external conditions that will impact your customer traffic.
Knowing the above is not so simple. Do you have immediate access to this information allowing you to make quick decisions without going to multiple resources?
In the customer service industry it is important to know your usual customer demand down to the 30 or 15 minute increment because your customers don’t want to wait more than 15 minutes to be served. Without a good handle on this your wait lines will get too long and potential customers will leave and go next door where the line is short.
The service industry’s challenge is access to the information needed for making timely decisions that lead to excellent customer service without overspending on wages.
1. How do I get ready access to my historical source data?
For restaurants, the best source of data is from your Point of Sales system. This data should feed into your staff scheduling software so that forecasts can be generated for you to schedule against.
ZUUS Dynamic Scheduling integrates with more than 40 POS systems.
For retail, the best source data is sales from your POS plus customer counts from your foot traffic monitoring software. It is critical to know how many people enter your store, not just how much people buy; if you don’t have enough staff on the floor to assist customers coming in you will miss out on revenue opportunities.
Now what do you do with the historical data to turn it into a forecast?
2. How do I figure out my sales growth index?
The most common practices to figure out your sales growth index are to either compare sales this year to last year, or a shorter timeframe comparison like, last months with this month. There are some more complicated models, which we will cover in another blog.
Example: Week 3 2016 sales were $5500, Week 3 2017 are $7300. Divide the sales for same time last year by the sales this year.
7300/ 5500= 1.3x
You have grown by 1.3 times. Hence to forecast for this year, times the same week last year by 1.3.
Keep in mind if you have changed your services, such as major menu changes (added breakfast) or expanded your store, last year’s comparison may not be a good indicator. Instead, some businesses use sales from 2 weeks prior and may add a small growth index if they are seeing week over week sales growth.
You will then apply the sales growth index to the historical sales source data to generate the right sales forecast for you coming schedule period.
3. Should I apply a rolling average (moving average method)?
A rolling average is where you look at a number of previous weeks, such as average the sales between 11am-11:30am on the last three Tuesdays instead of just looking at sales last Tuesday. The reason for doing this is to ‘smooth’ out the bumps, and reduce the impact of an unusually high or low sales period.
Be careful sometimes people average too many or too few weeks and they end up with an error that is higher than their increment. This is an indicator that you should not apply that average period. Experiment with averaging less or more weeks or look at a larger increment, perhaps 30 minutes instead of 15 minutes.
There is no one method that will work for all restaurants and retails stores. Factors such as growth levels, seasonal change on your customer demand, and school holidays have an impact on your customer traffic.
4. The first step of building your staff schedule should be to note what unusual factors might affect your customer traffic
Weather, local events, discount specials can all impact your customer demand. It is very useful to have a default method for forecasting. For example, use my POS sales from two weeks ago and apply a four week rolling average. However, think about what is going to happen in the coming week. Is there a school holiday? If so, it might make sense to base your source data off the last holiday period and then add a growth index rather than your normal default.
If you spend 5 minutes identifying variables that impact your customer traffic before you start to build your shift schedule, with the right scheduling tools you can create the right sales forecast, schedule staff to match this customer demand and you will save 5% labor cost percentage,increase profit and improve customer service – all in under 15 minutes a week!