Automated Sales Forecasting with Predictive Analytics – Making AI Real (Part 4)
In today’s organizations, the role of financial controlling or FP&A is not only to provide financial insights so business partners can make better decisions, but it is also to lead the way towards a more mature use of analytics technology including predictive analytics for sales forecasting.
Predictive Analytics – a Priority for FP&A
Moving up the analytics maturity curve from merely describing and reporting the past to gaining real insight and foresight into the future is a near-term priority for many financial leaders. Fifty percent of global fp&a teams are looking to implement predictive analytics by 2020*, and seventy-two percent rate “Predictive Forecasting and Planning” as either “very important or “important” for their company**.
Read more about the analytics maturity curve, including descriptive, predictive and prescriptive analytics, in Trending Technologies for BI & Financial Planning and Analysis Making AI Real (Part 2)
Predictive Analytics for Sales Forecasting
A great example for creating real value for the business with predictive analytics is predictive forecasting for sales. An accurate sales forecast is important because it drives many other business decisions. The sales forecast and sales budget are key inputs for setting the overall financial budget for the organization, including operational levers such as sales incentives, marketing budgets, product launches, new hires and so on.
However, sales forecasting is still a time-consuming process for sales planners, who are often reverting back to the good old Excel spreadsheet or other tools that often provide insufficient analytics and insights to inform the sales forecast for the next quarter, month or week. Advanced predictive analytics can help ease the burden on sales planners by automating rolling forecasts and providing executives with more transparency and smart decision support for enterprise performance management.
Addressing the Trust Question in Predictive Analytics
A major reason why organizations are still hesitating to implement more automated processes for sales forecasting with predictive analytics is simply a lack of trust and confidence in machine-generated results. Thirty-nine percent of planners see “Reliability of Results” as one of the biggest challenges for predictive planning and forecasting**.
One way to build trust and get buy-in from planners is to make the quality of predictions highly transparent for planners and managers alike. There are a couple of ways to do this. We at Jedox include the following functionality in Jedox AIssisted™ Planning, the integrated Jedox cloud service for predictive analytics and artificial intelligence:
A quality factor, so planners can transparently evaluate which prediction is the best for use in their forecast and can automatically select the best hierarchy level for the forecast
A comparison of predicted and actual sales over time, along with upper and lower bounds to determine the accuracy of the prediction and assess it against manual planning results
Screenshot: Sales forecast with quality factors displayed on the top left
With more confidence and a way for the user to assess the accuracy of predictions, planners start to feel more comfortable and increasingly rely on the guidance and recommendations provided by predictive analytics. This can lead to a higher quality of forecasts and a much faster time to completion, freeing-up valuable time and enabling a higher frequency of rolling forecasts or even a continuous forecasting system. Sales management gains higher levels of transparency and ability to recognize biases or risks by comparing the machine-generated recommendation for the best prediction to the human expert opinion.
Transforming Forecasting by Digitizing Human Expertise
The ultimate goal of predictive analytics for sales forecasting is to fully automate the forecasting process and enable continuous forecasting with real-time data. This is done by capturing and digitizing human expertise, essentially teaching a computer system to “think” like a human sales planner. This knowledge is then enriched with additional relevant data from inside and outside the organization. Of course, the upfront data preparation takes time, since the impact of various data sources and drivers, such as competitive information, market developments and trends, has to be analyzed and integrated into the model. But once the undertaking yields high-quality, automated rolling forecasts that can be shared seamlessly across the business, FP&A, sales planners and sales management will have made a significant step towards digital transformation. The business will be better equipped to detect early warning signs or leverage new opportunities. Now, FP&A and its business partners can spend less time on low-value tasks and more time on simulations that prepare the business for the future and increase business agility and adaptability.
Want to learn how Jedox customers are automating sales forecasting with predictive analytics? Watch our webinars “Spotlight on the Future – Predictive Analytics for Sales Forecasting and Planning” in English or German to hear about customers success stories and see a live demo.
*3 Steps to Determine How Financial Planning and Analysis Could Benefit From AI, Gartner, 2018
**The Planning Survey 18, BARC, 2018
You missed the chance to attend our AI webinar series? Keep calm and watch the recorded videos: