Most companies would welcome the so-called problem of having high-volume sales too robust to keep track of. In this case, Intel knew it was a good dilemma to have. However, it also recognized that it was losing out on potential revenue by not aligning its sales efforts more strategically.
The issue, according to Ivan Harrow, director of business analytics at Intel, was that after Intel sells components to Original Design Manufacturers (ODMs) and Original Equipment Manufacturers (OEMs) through its Technology Provider program, little data is available. And that means the sales organization doesn’t have enough data to efficiently support the reseller.
Intel works with more than 140,000 resellers (in addition to its core group of direct customers) who are members of the Intel Technology Provider program. These resellers (Intel’s customers) specify, design, build, and resell products based on Intel technology and solutions. The resellers purchase products through a network of Intel Authorized Distributors and then Intel’s sales organization provides online support, warranty services and marketing support. Intel employees around the world deliver a variety of different analytics and business intelligence solutions to distributors in various geographic markets.
When the Intel Technology Provider program started more than 10 years ago, Intel sold components to distribution, distribution sold to resellers and resellers built the final product to be sold to end users. Since that time, the market evolved and the trend toward smaller mobile devices shifted channel dynamics. Now larger ODMs and OEMs are building the end product and selling that product to distributors, who then sell it to resellers.
According to Harrow, “As the market shifted, we started to lose track of which products sold in which market and by which reseller, and it was hard to know what level of support to give each reseller.”
In the past, Intel focused on providing support to the largest accounts, but it became increasingly difficult to identify those accounts. The sales organization needed help determining which customers should receive the most support and the optimal time to contact customers during the buying cycle.
To solve the problem, Harrow and his team first set out to answer two questions: Which resellers should we focus on that have the biggest potential for high-volume sales to increase revenue? And, which products have the biggest relevance to that customer (for cross-sell and upsell opportunities)? The answers, he knew, would only come with the help of the sales organization, so a partnership was born.
Harrow’s business analytics team implemented a three-step approach. First, they worked to understand the business domain by analyzing how resellers engage with Intel and the sales organization. Then, they collected internal and external data to develop a predictive analytics engine. Finally, they optimized the engine, gathering feedback from the sales team and then working to continually improve the model over time.
The result was an engine that uses predictive algorithms and real-time data analysis to prioritize sales engagements with resellers showing the greatest potential for high-volume sales. The engine also recommends optimal contact time and proposes products to offer to increase cross-sell and upsell opportunities.
Next Harrow’s team explored which data sources they could tap into. They took registration information from their Technology Provider program using two data mining techniques: unsupervised clustering and supervised classification. They pulled in demographic information, training information and historical sales data, and used it to create customer profiles.
Then the team identified the revenue potential for each customer using purchased external data that contained additional customer information, including their recent investments and financing, and whether they were looking to expand into new markets. The team also looked at data regarding big buyers’ historical purchasing data and similarities with other products Intel sells.
At the beginning of the fourth quarter of 2012, Harrow’s team deployed a proof-of-concept in Intel’s Asia-Pacific online sales center. Just three months later, the solution yielded incremental revenue of $3 million. Harrow and his team took what they learned from that phase and applied it across other geographies, rolling out the engine in the Europe, Middle East, Africa, and North America regions. By the end of 2013, the company saw incremental revenue of $35 million over four quarters. By the end of 2014, it realized more than $76 million in revenue uplift.
Thanks to the predictive analytics engine, Harrow’s team was able to identify three times as many high-volume resellers in a faster time compared to what could be done manually. This information was then passed on to call center account managers so they could prioritize which customers to engage.
Harrow believes the predictive analytics engine he and his team created can be scaled to any other industry with the potential for sales, particularly companies that feel they have a potential customer base not being fully accessed.
“It’s critical to understand the data sources you have available, and if you don’t have them, you have to go out and find them,” Harrow says.
In addition, Harrow says it’s important to have a partner on the business side that will be your advocate.
In Intel’s case, the predictive analytics engine required a shift in how they do business. “It was much easier for a sales organization to hear from someone they know and trust on the business side, rather than hearing it from an IT man,” Harrow says.
Renee Morad is a freelance writer and editor based in New Jersey. Her work has appeared in The New York Times, Discovery News, Business Insider, Ozy.com, NPR, MainStreet.com, and other outlets. If you have a story you would like profiled, contact her at [email protected].
The IT Innovators series of articles is underwritten by Microsoft, and is editorially independent.
