How to Choose the Right Revenue Forecasting Models

Picture of Chris Fermoyle
Senior Business Consultant
It’s been a long week, month or maybe even financial year in the world of sales and there’s something on your mind that’s taking the unwieldy form of a revenue forecasting model.

In some ways, you can be sure that your current revenue forecasting model has paid off and your sales performance software is helping to bring in the deals. Most of your team are closing and justly earning their compensation while your territory expansions and resources are a good match for the products brought to market.

But still, there’s a lingering thread of doubt before you shut up shop for the weekend. In other words: ‘are we as a business really applying the most appropriate revenue forecasting model to support both our sales figures and sales teams and do I understand the basics of these very complicated formulae?’

Just ahead of that weekend wind-down, put your math head back on and reacquaint yourself with some familiar models and thinking.

1. Straight-line/Demand forecasting

The mother, father and obedient child of all methods, straight-line forecasting follows the simple premise of analyzing historic data and applying algorithms to more accurately predict future sales and trends. It’s tried, tested, respected and works. Mostly.

Is it the model for you? The output and reporting are particularly useful and clear for deciding on whether there is a case for territory expansion as well as helping to inform the entire supply chain on whether goods need to be reordered or reduced. It’s also popular for sales performance reviews of your team and objective setting as it relies on the value of what’s already been and gone. However, its success is dependent on very precise mathematical formulae and spreadsheets meaning the margin for error cannot be overlooked.

2. Moving Average

This technique looks at underlying data patterns to estimate future values and outcomes. Traditionally, these averages are calculated on 3- or 5-month figures, comparing the date sets within these to reach a median and predictive result.
Is it the model for you? This is a popular forecasting method for looking at short-term results within a very limited time-span–possibly gauging a sales reps commission earned over a quarter–but may not be so reliable when stacking up more historical records for scrutiny. There could also be good reason to doubt the figures if a fluctuation in the market–such as the current COVID-19 pandemic–is not specifically accounted for.

3. Simple Linear Regression

These models find relationships between 2 variables–1 dependent and the other independent. For example, the current size of your sales team in comparison to the launch of a new product or territory gateway. That’s putting something highly mathematical and complicated into the most simple of terms, which is why sales specialists are not always comfortable applying it, even though its forensic understanding of market forces and granular depth of company data may add rich and valuable layers to how various scenarios might be impacting sales, for better or worse.

Is it the model for you? The pause for action on applying this one is that while there may well be a correlation going on with 2 pieces of data, it’s not necessary that 1 causes the other circumstance to occur. Therefore, understanding the value of the relationship may not be obvious. However, get the right statisticians in the dashboard seat on this one and you can turn complicated formulae into sales gold.

4. Multiple Linear Regression

You can probably take a guess where this one is going by reading the one above. More impressive stats and data when applied correctly. Multiple linear regression is where more than 2 variables join in while there must be a single continuous variable. The clever bit is weighing up a host of projections which might be, recruitment budget, training/product knowledge budget and compensation plan outlay for a fiscal year. There can be many more variables across a whole lot of areas, but that static dependent one is your constant.

Is this the model for you? While introducing multiple fields and variables may sound like more of a headache than the simple model, the greater the amount of variables or factors you’re looking at, the more you can accurately determine when putting together a robust sales forecast.

Summary: The application of a sales revenue forecasting model may seem like a weekend migraine in the making, but the value of matching up with the right one cannot be underestimated. The correct formula will guide and drive business strategy, hiring decisions, as well as how you utilize your sales performance software.
If you want to find out how Varicent’s software and solutions can help drive your sales force to greater success, contact your rep today or visit:

Tags: Revenue Intelligence

Picture of Chris Fermoyle
Senior Business Consultant

You may also like