The holidays are upon us once again, and along with them, the corresponding annual shopping season. For dads like me with school-aged children, this is also the time of year when I turn to my favorite e-tailers to fulfill our family's wish lists. It's a tough call as to whether I love my children or hate the mall with more fervor, so in our house, “Santazon.com" with “Reindeer Prime" delivery has saved Christmas like the repentant Grinch at the end of the book. I know that it's certainly made me a lot merrier.
As we proceed to checkout with our shopping carts, we notice the e-tailer's recommendation engine making "frequently bought with" suggestions at the bottom of our screens. If I'm buying my 11-year-old son a new PlayStation 4 (Shhh...it's a secret), I have no problem with the e-tailer making the suggestion that he also perhaps could use an extra controller or might enjoy a specific game to go with it.
Most people would find that sort of suggestion unintrusive and helpful, especially if buying those things at the same time offers better value and makes the best use out of the money I'm already spending on his gift. Based on millions of purchases and data points, the concept of predictive analytics being used to make purchasing suggestions has become a natural and accepted part of the online shopping experience.
Using Predictive Analytics to Make Product Recommendations: What B2B Can Learn from E-Tailers
In the world of B2B sales, we also have the ability to make the most of our data points and engage in predictive upselling and cross-selling. At our last Technology Services World conference in Las Vegas, TSIA's own Jeremy DalleTezze and Judit Szilagyi talked about the different ways that technology companies can use similar principles to make relevant, timely product suggestions to their existing customers in the context of their interactions with the services team and support infrastructure.
Relatively uncomplicated data mining can identify which products and services are commonly purchased and used in conjunction with one another. The lists of purchased SKUs are analyzed, along with dates purchased and the account who made the purchases. The output is the creation of a set of rules which provide the estimated “win-rate” for potential new offers, given what the customer has already purchased. Product suggestions can be tied to common problems that customers may be having, syncing up with the products they've already purchased.
(Click image to enlarge.)
A slide from Jeremy and Judit's presentation that provides a visual map for the above explanation.
To use a fun example, if someone is buying graham crackers and marshmallows, predictive analytics can be used to identify that they're looking to make s'mores, allowing you to confidently recommend they buy chocolate bars.
How to Apply Predictive Analytics Data to Your Expand Selling Efforts
Obviously, this knowledge can help the product marketing team create valuable product and pricing bundles. Perhaps more compellingly, the outputs of this analysis can be key drivers toward Expand Selling, which is TSIA's systematic approach for safely using services touchpoints to generate leads, drive revenue growth, and provide better outcomes with existing customers. At several TSIA member companies, these analyses are being utilized in real time by their support services teams. In one leading company, scripts and talk tracks have been coded into their CRM, prompting support engineers to make upsell and cross-sell suggestions based on both previous purchases and common issues that the customer is having. By codifying and displaying these common fixes, the company now estimates that 24% of the “wins” generated by their team come from automated sources.
Many companies, including the one mentioned above, are also looking into integrating real-time product suggestions and the ability to make relevant purchases into their automated customer service systems. This technique, known as “machine learning” because the platform itself uses collected data to make the most relevant recommendations, takes the e-tailer model a step further into the B2B world. These automated product suggestions provide instantaneous results that aim to help the customer find an appropriate solution to their specific business problem, all with an incredibly low cost of completing the sale.
Not Ready for Predictive Analytics or Machine Learning Yet? Start Here.
Even if your company hasn't yet implemented machine learning or predictive analytics, you can still apply the same principles outlined here. As it turns out, you already have a tremendous database of product knowledge locked inside of the brains of your services and support teams. Chances are, they already anecdotally know which products, services and bundles work well together. Joining TSIA's Expand Selling discipline can give you frameworks and best practices for turning that knowledge into cost-effective upsells and cross-sells, including how to deal with training, lead qualification and tracking, compensation and more.
Most importantly, remember as you embark on this journey to keep in mind the Holiday spirit of helpfulness. TSIA stands by its mantra that “Helping will sell. Selling won't help,” and firmly believes that companies that keep this mindset can successfully tap into this incredible source of lead generation and high-margin revenue, while keeping themselves on Santa's "nice" list. Whether automated or human, these upsell and cross-sell recommendations must be provided in the spirit of helping the customer solve their business problems, achieve their desired outcomes, and maximize the value of their technology investment.