Predictive Analytics – It’s Just Getting Started once the Analysis is Done

Predictive analytics, along with social and mobile, is one of the hottest topics in the insurance industry today, and it’s becoming increasingly prevalent in strategies being developed.  While its potential is clear and results are proven, the thing about predictive analytics is that it’s much bigger than the “analysis” of the data.  It’s actually just one part of an entire ecosystem and strategy that must be in place to truly make it successful. 

Predictive analytics can be leveraged across the entire insurance lifecycle, ranging from sales and marketing to fraud, underwriting, claims and retention.  With the major industry focus on growth as of late, in particular profitable U/W growth, most insurers are starting to make predictive analytics investments in sales and marketing.  There are four major areas that must be addressed in any predictive analytics project, and I’m going to take a look at them from a sales and marketing perspective:

  • Context – in essence, predictive analysis is really all about context.  Without understanding the context of an interaction, a customer relationship, demographics, agency relationship etc., it’s difficult for predictive analytics to help make an accurate decision and recommendation.  The more information, the better.  Part of the context is understanding who the information is for and how will it be used.

    The obvious application of context is trying to understand the applicant’s or insured’s motivations.  Are they price conscious?  Have their needs changed recently?  Are they at risk of leaving, or ready to be offered an additional product?  Predictive analytics isn’t just for use with insureds or applicants, it’s also meant to be applied to agents, both career/captive and independent.  Predictive analysis can be applied to producers (agents and brokers) to make them more effective, and is critical for insures who support mostly independent sales channels.  Insurers need to understand: is the producer young and hungry or a farmer gliding to retirement?  Does the producer respond well to an insurer helping them sell within their book of business, or do they find it invasive?  Can we identify which producers are actively selling or who is blocking the market or who needs some help to be successful? 

  • Business Strategy – most insurers are looking to predictive analytics to predict “what the customer will do, what they will respond to and what will be the best path forward with the client”.  The real question is, “what” are you trying to predict?  Predictive analysis can only have a meaningful impact if it’s applied to a well-thought out and articulated business strategy.  Predictive analytics must be tuned to support a strategy that can both be effectively executed and measured.   As insurers look to enter new markets, seek out specific client segments, develop their channel strategies, predictive analytics offers an invaluable tool to effectively reach strategic goals faster than ever before.
  • Process – In order for predictive analytics to be truly effective, there must be an adequate process model, preferably as automated as possible, to take the information from the predictive analytics and “do something with it”.  Insurers need to foster both a business and IT environment that can leverage the benefits that predictive analytics brings to the table.  As predictive analytics strategies are developed, the surrounding environment needs to be flexible enough to be able to respond and adapt appropriately based on business goals and objectives.
  • Execution – This is the tip of the spear.  How does the predictive analytic results, business strategy and associated processes get delivered into the real world make a difference?  Is the customer interacting with an on-line quote/sales engine?  Or is an agent working with a client?  Is the agent a career/captive agent or independent?  Preparing channels for executing in conjunction with predictive analysis will take communications, awareness and possibly some training.  Not preparing for executing on a predictive analytics strategy is wasting opportunities for success. 

Often, when there is a lot of hype around a new technology, companies will jump in with both feet and not understand what they’re getting into, or why – but feel they need to because “it’s important”.  It reminds me of the days when every new release of Microsoft Windows was greeted with a media frenzy and long lines of people waiting overnight to be the first ones with the new release.  A reporter was interviewing people in New York at 3:00 AM who were waiting in line overnight for the latest Microsoft Windows launch the next day.  The reporter interviewed one of the people in line who felt that he was in on something big and couldn’t wait to buy his copy.  When asked the follow-up question about what was so great about the new product and how he would use it, the customer replied “… to be honest, I have no idea, I don’t even own a computer, it just seems to be too important to pass up”.  Don’t be that guy. 

If you’re interested in hearing more on this, join me on July 18, when I’ll be teaming up with Analyst Chuck Johnston of Celent to host a live webcast sponsored by Insurance Networking News.  The title of the complimentary webinar is, “Predictive Analytics for Insurance: Hype or Holy Grail?”  You can register for it here: