psychic
PHOTO: clairewinterphotography

Getting sales right is a tricky business. Gaining customers' attention requires a coordinated effort between marketing teams, who guide customers along their journey with relevant messaging, and sales teams, who interact with those same customers during their journey. 

While we hear so much about marketing teams using analytics to plan multichannel experiences, data can also help sales teams improve the accuracy of their sales forecasts, allowing them to better align resources in areas which have the greatest impact. 

Building a More Realistic Sales Forecast

Sales teams have traditionally relied on past performance to guide their forecasts for sales targets. To imagine how a forecast is set, imagine a sales top performer who hits above 100% of his or her sales target. Most sales managers will review weekly or monthly sales, then see how to sustain that sales rate. The manager may be tempted to set that top performer’s target for the next sales period at that rate or a multiple. You can also imagine a lower-tier sales performer achieving only 90% of the target. That person can receive the same target for the following year. This forecast approach has been standard practice among managers in many industries.   

But a good statistical analysis can better determine if a higher sales volume from a recent sales period warrants further investment to support sales. Without it, managers can misjudge a temporary spike as a longer-term trend. This kind of mistake can lead them to set bad sales targets for the sales team, expecting them to chase down sales growth when no opportunity really exists. Imagine overworked top performers tempted to leave your firm because the sales targets created unrealistic expectations of job performance.  

A statistical analysis can better support a manager’s intuition on how long a sales moment could last. Forming a prediction model based on sales data can be done easily these days. 

Related Article: How to Deliver Credible Marketing Pipeline Forecasts

How to Create a Prediction Model With Sales Data

For example, sales history for a salesperson’s territory can be imported into an advanced solution such as SPSS or into a dedicated programming language like R or Python. The data can then be treated as a time series. You can then apply a basic statistical analysis, such as means of sales volume or range of sales. Examine the sales data to establish a baseline trend and then see if that trend is changing significantly. Your analysis can also determine how to treat your sales history within a predictive model. Sophisticated techniques, like Bayesian Times Series, and frameworks such as Facebook Prophet can then forecast how long a meaningful sales trend will exist. Managers can then use that predicted time to better decide how long to allocate budget and sales team resources.  

No matter which approach you take, when a sales manager determines a sustainable sales trend, it will help them set more realistic sales targets. It can also improve the quality of conversations with the marketing team. As marketing campaigns have become multichannel, bringing sales into the discussion will help sales teams align their interactions with customers to the messages marketers are delivering at every step of the multichannel experiences. Imagine tailoring sales technique to customers in the consideration stages — think of the increase in conversions that a coordinated campaign between sales and marketing can create. 

Marketers are learning how data empowers them. Using data to improve sales forecasts is the simplest first step in empowering sales teams. Sales managers have an opportunity to receive the same data-savvy education marketer managers have for identifying good decisions and smarter risks.  

Related Article: How Marketers Can Get Started With Time Series Data in R Programming