Should Banks be Happy with a Silver Medal? (Part Two)

In Part One of this post, I described how the current “crisis in confidence” for banks may actually have a silver lining for progressive organizations who implement a Next-Best-Action (N-B-A) marketing approach.  The three key components to an N-B-A solution are analytics, real-time execution and a central decisioning hub.

 

Analytics

For those unfamiliar with this term, and to do a level-set, let’s start with a definition:

Predictive Analytics encompasses a variety of statistical techniques that analyze current and historical facts to make predictions about future events.  In business, predictive models exploit patterns found in historical and transactional data to identify risks and opportunities. Models capture relationships among many factors to allow assessment of risk or potential associated with a particular set of conditions, guiding decision making for candidate transactions.(1)”

OK, so why does this really matter?  There are two aspects at play that make the use of predictive analytics compelling:  product mastery and data growth.  While a typical bank may have over 100 products, employees usually master less than a dozen.  These are their “go to” products with reasonable success rates that they can cover confidently.  Your best staff may exceed that average, but with turnover rates from 25-50% depending on role, how many of your “best” people are around?  Also, is your bank truly building long-term value if your employees commonly sell what they “know” and not the best match for the customer?  Self-service isn’t a silver bullet either.  The complexity of some products limits unassisted channel uptake and drives traffic back to agent assisted channels. 

There is also an exponential growth in data associated with this wonderfully wired world we inhabit.  According to IDC (2)[CB1] , the amount of data transmitted worldwide in 2012 is predicted to be 2.7 zettabytes (1 ZB = 1 Billion terabytes or TB) which is a 48% increase over 2011.  While that represents ALL traffic on the planet, even data at most banks are in the hundreds of TBs with people now talking about petabytes (just 1,000 TB).  Predictive analytics can yield a treasure trove of insight using all that data to identify what products a customer may be interested in or their likelihood to attrite.  Models, derived statistically, will consistently outperform seemingly rational intuition, experience or “gut” insights in the face of that growing data tsunami. 

With recent advances in predictive technologies, a PhD in statistics is no longer required to understand the business purpose of the analytics.  Traditional attribute-driven segmentation can only tell marketers how to sell to a homogeneous group, it cannot describe exactly who to sell a specific product.  With behavioral segmentation, an individual can become a “segment of one” in which non-linear interplay between individual dimensions can be used to select exactly the right action or offer for exactly the right individual.  

However, it’s not clear all banks are utilizing this approach to its fullest potential.  This spring, BAI, in conjunction with Pegasystems, conducted a survey (3) polling more than 200 retail banking executives on customer experience – defined as both sales and service interactions.  We asked respondents how often they refreshed their sales and marketing analytical models.  50% of respondents either don’t use any analytical tools or just update them annually. 

That “annual” response exposes the one weakness of predictive analytics – staleness.   Because models are based on a “snapshot” in time, they lose predictability with age and the market will change before the model is updated.  The true customer behavior is constantly shifting along with:

  • Consumer sentiment and confidence
  • Governments, economics and regulations,
  • Competition and product offerings, and
  • Technology (with a nod to the new iPhone 5)

While everyone wants to put a stake in the ground, I’d argue that we still have not reached a “new normal” yet after the Great Recession.  So without hiring an entire graduating class of statisticians, you can counteract this affect by including self-learning or adaptive models into your N-B-A strategies.  Adaptive models update automatically after each customer response and are therefore, always up to date.  Adaptive models that recognize these “zigs and zags” of customer behavior can then reinforce your predictive model(s) in champion/challenger settings to provide a critical safety net for your marketing efforts.

With that practical use of predictive analytics established, I’ll move on to real-time execution and a central decisioning hub in Part 3.  Additionally, I will be hosting a webinar covering all these topics and more with Tim Crick, formerly of Royal Bank of Scotland - Head of Customer Contact Architecture.  Please join us Wednesday, September 19th at 2:00PM EDT for

 

2012: A Marketing Odyssey

The Customer Experience Journey with Next-Best-Action Marketing

http://www.bai.org/bai-events/EventDetails.aspx?ec=0564

 

References

1 - http://en.wikipedia.org/wiki/Predictive_analytics

2 - http://www.idc.com/getdoc.jsp?containerId=prUS23177411

3 - https://www.pega.com/resources/bai-survey-insights-into-todays-customer-service-and-sales-experience