Predictive Analytics Times
Predictive Analytics Times
EXCLUSIVE HIGHLIGHTS
SHARE THIS:

4 months ago
Wise Practitioner – Predictive Analytics Interview Series: Dean Abbott at SmarterHQ

 

By: Eric Siegel, Founder, Predictive Analytics World for Business


In anticipation of his upcoming conference presentation at Predictive Analytics World for Business Las Vegas, June 16-20, 2019, we asked Dean Abbott, Co-Founder and Chief Data Scientist at SmarterHQ, a few questions about their deployment of predictive analytics. Catch a glimpse of his presentation, Case Study: Improving Customer “Lifetime” Value Predictions, and see what’s in store at the PAW Business conference in Las Vegas.

Q: In your work with predictive analytics, what behavior or outcome do your models predict?

A: In my current work, most of our models predicts shopper/customer behavior. The models are binary classification models that predict likelihood to purchase within a fixed time period, most commonly 7 days. We call these “purchase propensity” models.

Q: How does predictive analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?

A:  The purchase propensity models are used in several ways. For some clients, the purchase propensity models identify customer segments to contact with a message. One creative use of these segments is to identify non-obvious shoppers poised to convert; the obvious ones are easy to find using the expertise of marketers to identify the good shoppers. The models identify those that may have been missed. A second use case is to create a customer segment and give them an offer, such as free shipping or a discount, except for those who have very high likelihood to purchase; just let these shoppers do their thing!

Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: I can’t provide specific results for active campaigns. Instead of creating deciles of scores, we create unequal sized bins for campaigns: 99th-100th percentile, 95th-99th percentile, 75th-95th percentile, and others for smaller scores. For some clients, the top 1% of scores have a lift of 50x to over 100x compared to the average response rate.

Q: What surprising discovery or insight have you unearthed in your data?

A: Purchase propensity models identify several different kinds of patterns in shopper behavior depending on the type of business. For most retailers, purchase and visit recency are very important predictors of near-term future purchase behavior; the best customers buy over and over again. However, some shoppers are highly likely to purchase in the near-term but aren’t very engaged on the website. These shoppers don’t do comparative shopping: they know what they want, cart, and purchase.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: The talk I will be giving describes a new model we are including at SmarterHQ: customer lifetime value (CLV) models. The talk will compare traditional CLV models, based on historic transactions only, and CLV models that include online behavioral information as well. We will quantify if behavioral data improves CLV models, and if so, by how much.

—————————–

Don’t miss Dean’s presentation, Case Study: Improving Customer “Lifetime” Value Predictions, at PAW Business on Tuesday, June 18, 2019 from 10:30 to 11:15 AM. Click here to register for attendance.

By: Eric Siegel, Founder, Predictive Analytics World for Business

Leave a Reply

Pin It on Pinterest

Share This