Accelerating analytic learning—it’s one of the top challenges my team is working to solve with many FICO clients. No wonder it’s been a running theme on this blog, and of my own posts.Our most successful clients are accelerating analytic learning by pushing the boundaries beyond traditional champion/challenger testing. Here’s how:
Conducting designed experiments, whose results can be accurately analyzed and causes of variations understood. A well-designed series of experiments tests carefully to explore as well as confirm key action-effect relationship hypotheses. In other words, the relationship between how the bank treats their customers and the consequences of such treatments that drive the P&L metrics for the bank. This is a powerful framework for learning that complements the traditions of “champion” vs. “challenger” testing.
Probing beyond the edges of business as usual (BAU). An action-effect framework for creating experiments is a more surgical way of testing and learning. By mathematically tying the P&L drivers...
My last few posts have discussed the analytic learning loop. By accelerating feedback about market performance, analytic learning loops enable banks to target the right products to consumers most likely to respond and generate profits. And banks can adjust decision strategies to boost results while campaigns are still ongoing.So what does this approach look like?The fundamental component of an analytic learning loop is an analytic learning hub, through which insights from data analysis and learning from production testing flow. This hub will generally be comprised of analytic data marts (continually refreshed from internal and external data sources), a repository for past treatments, a variety of analytics to monitor what is happening as well as project what might yet happen, and a methodology for creating actionable diagnostic reports. As shown in the graphic, the learning loop is then "threaded" through both acquisitions and originations processes. This "threading" occurs through shared access to the hub as well as automated data feeds from...
These days, economic volatility, new regulations and competitive forces all make analytic learning speed and yield critical to success. In my last post, I discussed how an “analytic learning loop” accelerates feedback about market performance, enabling banks to make decisions based on how consumers are behaving today. By providing a systematic approach to learning from recent results, analytic learning loops improve a bank's ability to:
Design innovative products customers need, based not on their past behavior, but on how they're responding to and using your products right now.
Deliver more effective pre-approved lead lists. Richer customer profiles enable more granular and insightful segmentation of prospect populations. Higher-performing lead lists boost the performance of marketing campaigns, and the productivity at call centers and branch offices.
Improve response rates and early-life account performance by pinpointing individuals who are not only most likely to accept a specific offer, but to use the product in a manner that is profitable...
These days, portfolio growth and a bank's ability to quickly understand and adapt to changing consumer behavior are interconnected. The challenge, of course, is that traditional approaches to acquisitions and originations, based on analyzing past behavioral data, provide a historic view of customer risk and reward potential. But in today's markets, customer behavior is changing. The situation is something like this familiar scene in films and TV: A guard watching a live video monitor is fooled when a segment of video is copied and looped in place of the live feed. The monitor shows the scene unchanged, even though there's actually a lot of activity going on. Instead of this fixed, historic view, banks need growth strategies to reflect what's going on in markets now. The way to achieve this without organizational and infrastructure upheaval is to thread an “analytic learning loop” through existing processes. An analytic learning loop provides timely, constant feedback on the performance of decision strategies in operations, and early warnings of...
What does the “new normal” of consumer behavior look like in the wake of the economic crisis, and what are the effects on consumer credit risk? It’s a subject I was invited to discuss by the Kellogg School of Management at Northwestern University at their recent Risk Summit. Other speakers included execs from Cap One and Discover. The topics we discussed included the phenomenon of strategic defaults (and how banks can get a handle on this problem); the need to understand your customer better; and the role of data, analytics and business expertise as banks begin to once again loosen credit risk in their lending portfolios. It was a great conference that sparked some lively debate. Take a look at the brief article the folks at Kellogg have published on their website that summarizes the event. They even posted a short video interview with yours truly.
