Forecasting customers is an essential part of any business. It allows you to predict future sales, allocate resources effectively, and make better decisions about your products and services. In this blog post, we will discuss the basics of forecasting customers and provide tips for improving your predictions.
Why It Matters
Forecasts are all about the future. It’s impossible to overstate how essential it is for a business to have an accurate sales prediction. When leaders can trust forecasts, privately-held businesses gain trust in their operations. In addition, accurate forecasts help to establish the credibility of publicly traded businesses.
Forecasting customers is important for several reasons. First, it allows you to track your current performance and spot trends. This information can help you make informed decisions about where your business is headed and what changes (if any) need to be made. Second, forecasting helps you prepare for potential growth. You can use your forecast to identify opportunities and plan for how you will handle increased (or decreased) demand. Finally, a reliable forecast can be used for scenario planning and stress testing your business with different operational and product decisions.
Breaking It Down
When you work on any forecast, you need to step back and break the forecast down into pieces. What are the drivers behind the forecast?
For customer forecasting, these drivers will be things like historical demand, market capture, and seasonality depending on the forecasting method you choose. Once you have identified the drivers, you need to think about how they will change over time. For example, if you run a promotion, your customer traffic should increase in the short term.
While there are many qualitative demand forecasting methods, those methods tend to be in the domain of sales and marketing. We will focus on quantitative forecasting methods.
- Test Marketing – Sell products in one specific market and extrapolate that demand to similar markets
- Time Series Analysis – Use historical activity, including seasonality, to predict future results
- Regression Analysis – Use a variable that you know to forecast a variable that you don’t know
- Econometric Models – Replacement demand + new owner demand = sales
We will use time series analysis since it is most effective for the average business.
The Forecasting Process
Step 1: Determine What You Need to Forecast
We’ll need to predict the following components in order to produce a reliable customer forecast:
- Customer Demand
- Favorable / Unfavorable business changes
Step 2: Collect Inputs and Assumptions
For each of the components from step 1, you will need to collect the inputs and assumptions behind them. For example, business changes may be driven by promotions that you could get from the sales or marketing team.
If you are a new business, you may need to estimate things like customer demand or seasonality based on competitors. If you are an existing business, you might want to consider historical data.
Step 3: Layout the Forecast Model
Once you have determined what to forecast and collected the inputs, it is time to start building a model. Forecasting models can be as simple or as complex as you want them to be. For example, if we’re forecasting customers for one store with many years of historical data, the spreadsheet for your model would be really simple.
On the other hand, if you are forecasting customers for 1,000 stores across a country, you will need a complex model with more details.
Step 4: Run and Adjust the Forecast
Once your model is set up, you simply need to run it and adjust the inputs as needed. You will want to do this on a regular basis, especially if your business is growing or changing. Forecasting is not a one-time event; it should be done regularly to ensure that your numbers are accurate.
Step 5: Review and Summarize
You should look at your forecast results to ensure they make sense and summarize them in a way that works for your business. For example, if you are forecasting customers on a monthly basis, you may want to do quarterly and annual views. If you are forecasting by store, you may want to view by region. You may even want to look at trend views to ensure there aren’t any outliers.
Let’s Walk Through An Example
Forecast Customers For A Restaurant
Let’s walk through a basic customer forecast for a restaurant. We are going to use time series analysis which is a really great way of using historical data to predict the future.
First, let’s lay out our forecast model and the historical data that we have on hand. As mentioned above, we want to use the following items to forecast (in addition to historical data):
-Customer Demand: These are the actual customers who showed up each day during the historical period
-Seasonality: In this case, how the day of the week impacts customer volume
-Capacity: The maximum capacity of the restaurant on a given day, no matter the demand, customer counts can’t exceed the capacity
-Favorable / Unfavorable business changes: Promos, special events, or holidays that could impact the business
Here is what the forecast model looks like:
Now that we have historicals populated, we can turn our attention to the forecast periods.
First, let’s forecast the drivers.
-Customer Demand: This will be the average demand over the historical period
-Seasonality: This will be an adjustment for the seasonality factor for that specific day of the week over the historical period
-Capacity: The customer demand will be capped at this capacity. Usually only a fundamental change to the restaurant (expansion, outdoor dining) or operations (opening hours) can change this number
-Favorable / Unfavorable business changes: We will assume a buy one get one half off promo is launched for 3/25/22 to 3/29/22
Once the drivers are built, we will load in the formula. Take the assumed demand, multiply it by the seasonality and any changes. Then, use an IF formula to restrict the maximum demand to the capacity.
Here is our final forecast:
Tips and Tricks
Finding Data And Assumptions
Forecasts are only as good as the data and assumptions you put into them. So where can you find solid data? If you have an existing business, the first place to look is your financial system. Historical data is one of the best inputs to a forecast. You can also work with your operations teams to understand what it will take to deliver a certain level of performance.
For a new business without historical info, you will have to dig a bit deeper. Economic data and market research are your best bets. This can include digging into resources like the Consumer Product Index (CPI) for inflation or studying your competitors.
Step Back And Do A Gut Check
As you get into the weeds of your forecast, it is important to step back and ask yourself, “Does this make sense?” Think about how the forecast looks year-over-year and sequentially. Do you have the capacity and workforce to even deliver the forecast? Do the trends seem reasonable or are there unusual outliers in the forecast?
It is critical to sanity-check your work and ensures you put out a great product
Build For the Future
When working on a forecast, do yourself a huge favor and build it for the future. Well, obviously a forecast is for the future, but I mean the model itself. If you are running a forecast today, you are likely to run the forecast again. Make sure the model is dynamic enough to pull in new actuals and roll forward for future time periods. Avoid hardcoding, and try to link everything up to data tables. Make it clear which periods and cells are actuals and which are forecast.
This may take some extra time to set up, but it will really pay off down the road.
Forecasting customers is a critical component of any business. The process of forecasting involves estimating future revenue by predicting the number of products or services a sales unit will sell in the next week, month, quarter, or year. Forecasting models can be as simple or as complex as you want them to be, and you should review and adjust your forecast regularly to ensure that your numbers are accurate.
When forecasting customers, be sure to use solid data and assumptions, and do a gut check to make sure the forecast makes sense. Finally, build your model for the future by making it dynamic and easy to update. Forecasting is an essential tool for any business, so take the time to do it right.
Have any questions? Are there other topics you would like us to cover? Leave a comment below and let us know! Make sure to subscribe to our Newsletter to receive exclusive financial news right to your inbox.