By: Eric Siegel, Founder, Predictive Analytics World
In anticipation of his upcoming conference presentation, B2B Tech 4sight: Identifying the Right Product Solution for the Right End Customer, from the Right Reseller at the Right Time, at Predictive Analytics World for Business Las Vegas, June 3-7, 2018, we asked Olaf Menzer, Predictive Data 
Analyst at Ingram Micro, a few questions about his work in predictive analytics.
Q: In your work with predictive analytics, what behavior or outcome do your models predict?
A: The models we build in the Global Analytics team predict purchasing propensity at several levels of the B2B tech distribution supply chain. For example, we calculate propensity for different product categories, reseller types, and end consumer verticals (e.g., manufacturing, health care, education, construction), often in terms of probabilities under a given a time frame (usually 1 week, 1 month or 1 quarter):
- Reseller Level: Lead generation, Customer growth incentives
- End Customer (Consumer) Level: Propensity to purchase and propensity to refresh
- Product Level: Cross-selling and up-selling opportunities
Q: How does predictive analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?
A: At a high level, by delivering pricing optimization and R&D solutions internally, as well as designing and executing marketing analytics programs offered to external clients (e.g., technology vendors and resellers). We also enable data-driven decision making throughout business verticals in organization, for example Vendor Management, Purchasing and Transportation. Finally, we provide value through data aggregation, synthesis and visualization, for example with dashboards and interactive charts – in an enterprise of our size there is a large value in capturing information from different parts of the business, enriching it with external data sources and connecting the dots.
Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?
A: The ROI of the predictive analytics programs (e.g., lead generation, cross-selling) that we are offering to vendors can be in excess of 20x in terms of incremental revenue, depending on vendor size and product category. We note that we have the ability to predict consumer behavior and provide estimates of returns prior to running a campaign.
Q: What surprising discovery or insight have you unearthed in your data?
A: If the sales volume of a technology product or segment through Ingram Micro reaches a certain threshold, we’re finding a very high correlation with trends in the larger market as a whole. This enables us to capture early and ongoing signals and to create detection mechanisms of high purchasing propensity for trending products and the most popular IT solutions, and to match resellers with end consumers. For example, we can estimate purchasing propensity for newly introduced products as well as hundreds of different established product categories within virtually any end customer vertical or industry.
Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.
A: “Data first, models second.”
What is meant by this is that any meaningful predictive outcome needs to be based on capturing the right business facts and numbers, and then enriching that information with context and expertise quantitatively. Usually, it takes variable transformations and grouping, resampling, feature engineering and labeling processes in the input space to make our models more accurate, to drive business value and high ROI.
———————
Don’t miss Olaf’s conference presentation, B2B Tech 4sight: Identifying the Right Product Solution for the Right End Customer, from the Right Reseller at the Right Time, on Wednesday, June 6, 2018 from 2:15 to 3:00 pm at Predictive Analytics World Las Vegas, 2018. Click here to register to attend.
By: Eric Siegel, Founder, Predictive Analytics World