How Prescriptive Analytics Puts Complex Decision-Making on Rails

In this special guest feature, Nick Hedges, President and CEO at Velocify, discusses the importance of prescriptive analytics and utilizing data effectively, particularly for B2B sales. Nick is a 15 year veteran of the Internet and software as a service (SaaS) industry. He joined Velocify in 2008 as SVP of business development and then held various roles at the company, including head of sales and chief revenue officer, prior to becoming CEO in 2011. Prior to Velocify, Nick was a case team leader at Bain and Company, where he led strategic assessments in the technology, consumer products, media, and private equity industries. Earlier in his career, Nick worked at Andersen Consulting (now Accenture) on Internet and process-reengineering projects and at Ogilvy and Mather Advertising as an account manager for big brands, such as Kodak and Ford Motor Company. A Fulbright Scholar, Nick holds an MBA with Distinction from Harvard Business School and a bachelor’s degree with first class honors from Manchester University.

The amount of data and information available today is increasing the complexity and pace of business, and companies are struggling to maintain control at high velocity. The shift has affected all aspects of the organization, from marketing and sales to product development to customer relations, as expectations for turnaround times get shorter and shorter.

The Big Data revolution seems to be everyone’s answer to this macro problem. But turning all the world’s data into useable and actionable information quickly has been one of the top challenges since the term Big Data came onto the scene about eight years ago.

Will 2016 finally be the year where we start to see data driving truly intelligent processes, in an automated, transparent, and scalable way? Prescriptive analytics could provide some answers.

What Is Prescriptive Analytics and How Does It Work?

Prescriptive analytics is often considered the third wave of the Big Data revolution, coming after descriptive analytics and predictive analytics.

Descriptive analytics – addresses the first step in evaluating massive datasets to give users a clearer picture of what their data means by uncovering patterns and representing meaningful information through intuitive charts and graphs.

Predictive analytics – moves the process a step further by taking that data and running additional analysis (often with comparisons against historical datasets) to forecast future trends and behaviors. For example, it could predict how many customers are likely to terminate their contracts or what percentage of interested buyers are likely to convert in a given time period.

Prescriptive analytics – adds a third layer to this technology by evaluating possible actions in response to the data and determining which is most likely to produce a desired outcome. For example, a prescriptive analytics algorithm can tell a salesperson which lead in the company pipeline to follow up with next (and what means of communication will be most effective) based on a complex analysis of different factors ranging from demographic data to situational context as well as how similar leads reacted to different actions in the past.

Why Transparency and Control Are Required to Achieve True Sales Velocity

It’s clear that shorter business cycles mean that companies need to react faster than ever, but that doesn’t mean that complete control should be handed over to a technology or algorithm, i.e. a predictive model. In fact, when sales acceleration providers talk about factoring in all sorts of extraneous data – like cycles of the moon and sports scores – without showing how this helps the seller, it makes it harder to bring sellers on board.

But the combination of predictive and prescriptive analytics can prove enormously helpful for enterprises that need to do more in less time.

Transparency is critical to user adoption – Consider Waze and Siri for an example of the importance of transparency for software that gives advice. Waze experienced massive user growth with minimal marketing because of its transparency: all the data it’s using to make its predictions shows up on the map as you drive. Meanwhile, Siri was heavily marketed but ended up as the butt of many jokes when it couldn’t make good on those early promises. Although Siri has improved to become much smarter and more effective, it’s still struggling to regain users who were turned off by the “black box” approach that made it hard for people to see how the technology evolved in response to its early stumbles.

Providers of sales acceleration technologies – or any other enterprise software, to be honest – must pay attention to this lesson. With automatic lead prioritization, for example, technology should not just tell salespeople the next step to take in the process, but offer a transparent view that shows the seller why this precise action is important now.

A certain level of human control is needed – Continuing with the example of Waze, consider for a moment that you are looking for the fastest route from work to home. You put the data into your phone and Waze gives you the recommended route. This is one piece of data, but there might be other things important to you on your journey home. Maybe you need to make a stop at the grocery store for milk, or maybe you prefer taking a scenic route for the enjoyment. There is some level of human control needed to ultimately drive the best result for you.

Turning this back to sales, a simple predictive score gives you an indication of product fit based on key attributes, but there may be other factors worth considering. Combining prescriptive and predictive analytics adds more factors into the equation, such as the responsiveness of a prospect to your outreach. Have you tried calling them three or four times already with no response, and if so is that prospect more or less important than a new lead that may not appear to be the perfect product fit? A comprehensive prescriptive tool truly makes the predictive score more actionable.

The bottom line is that “black-box” prescriptive analytics solutions, which simply deliver a command without providing additional background or context, can alienate users, especially if the recommendations don’t quickly translate into improved results. Effective prescriptive analytics must deliver more transparency and control, allowing admins to drive the prescriptive rules and workflow, and users to understand what goes into a recommendation, which establishes a greater level of trust.

Sign up for the free insideBIGDATA newsletter.