Blog Posts By: Roman Lenzen

Now that the New Year is upon us it is time to review analytics in 2011 and look ahead to 2012.

2011 in review:

• Analytics became a Sport: Several analytic competition sites became highly visible in 2011, the most prominent being Kaggle.com.  These sites host analytic competitions that anyone can enter and provide a rich learning environment for both new and seasoned analytic professionals.  The winners are not just those who place in the money - the winners are everyone who participates and learns within the community environment.   

• Moneyball and Analytics Hype:  Speaking of sports, I'm a baseball fan so I naturally enjoyed the movie Moneyball.  Analytics hype continued to climb in 2011 with the help of that movie.  See this recent post for more on my thoughts on the state of analytics hype in 2011.   

• 2011 was the year of R (an open sourced analytic solution widely regarded as one of the most powerful analytic tools available): Corporate adoption gained momentum in 2011, as the corporate analytic community continued to understand the power of the solution and myths such as it is not a tool for production environments were quickly debunked.
  • 50% of the Kaggle.com winners used R

Customer Analytics seems to be one of the hottest topics in business today, with the newly coined 'Data Scientist' role in high demand.  And with all hot topics there is hype and reality.   The reality is that analytics can provide a big impact on bottom line profits.  The hype is that it can provide a significant ROI in all environments.  In most environments, you need three critical components for customer analytics to provide a significant impact:

In a previous post I laid out the process to create the collections database and initial collections analytics during acquisition.  This post will focus on how to utilize the database and analytics within proactive communications in order to optimize collection efforts then the ongoing tracking and enhancements required. 

1.       Treatment Strategies: Analytics alone will not enhance collections efforts.  The predictive analytics and segmentation systems must be converted into actionable treatment strategies which will 1)reduce the cost of collections and 2)ultimately, minimize the outstanding debt through increased collection payments.  Treatment examples which need to be setup and tied to each individual collections customers include (again, these are examples)

In a recent post I provided thoughts on how to bridge the actionable analytics gap assuming the analytic skills, tools and resource gaps are fulfilled and yet a large analytic divide still may exist in your organization.  In this post I would like to go back and address analytic skills, resources and tools.  I'm not going to provide suggested solutions, as all environments differ, but, I would like to provide key considerations as there are many analytic buzz words and hype in the marketplace, some of which are valid and some should be hampered. 

Analytic resources require several skills including database, technical and analytic skills as well as a strong business acumen.   A few key considerations to keep in mind-

Customer insights are best delivered in a visual fashion, so the business constituencies may easily digest the information in terms they understand.  To enhance your visualizations, and bring them up to warp speed, a time dimension should be added.  Adding a time dimension to your insights allows your audience -

Deriving insight from historical customer and prospect interactions (online and offline) has several implications, many of which can positively affect bottom line revenue or profit if understood and implemented correctly.  Presenting the insight in a form so that the business constituencies may easily digest the information, and start to visualize how it may be implemented to effect bottom line revenue (or profit), is one of the biggest barriers in insight and analytics adoption today.  What are the primary issues in bridging that analytics gap?  Skills, tools or resources?   Yes, you need the right tools and resources, with the right skills, but those are table-stakes and many institutions are still not bridging the analytic divide well with all three in place. 

A Health Care company and a data mining competition consortium have teamed up to offer the largest data mining prize in history and anyone can participate.  The primary goal of the Heritage Health Prize is to use de-identified patient data to predict future hospital stays which then may be used to better forecast stays, and those associated services, as well as to develop enhanced prescriptive actions which will reduce future stays and costs. 

Online ad spend is predicted to reach $100 Billion in 5 years, up from $61 Billion in 2010.  Who is best poised to tap into that incremental online advertising revenue, other than the advertisers?

Paid search aside, which is dominated by the behemoths (Google & Yahoo), it will be those firms who can store online data, analyze it quickly and make the insights actionable within all channels in order to make the advertising more relevant to the consumer. 

Now that we rang in the New Year, I thought we should take a look back on a few major 2010 analytic landmarks and then look ahead into 2011.

2010 in Review

Consolidation: Analytic consolidation continued to be highlighted by IBM's purchase of Unica, Netezza and Coremetrics.  Overkill especially given IBM's purchase of SPSS in 2009?  Maybe not given the heightened focus on online, big data and analytics.

"When I can't see clearly, I know you're still there.

When darkness swarms around me, I know you're still there."

When it comes to online footprints we are starting to see the light and it's old technology that is slowly illuminating those footprints and tying the steps together effectively so analytic consultants like myself may convert that data into actionable insights.

In Customer Analytics > Optimizing Your Collection Efforts, Part1  I focused on how companies are enhancing their collection efforts by setting up risk treatments at the point of acquisition.  Now that the customers are on board, and a portion of the customers have entered into collections, what can be done to optimize collection efforts and thus minimize the amount of outstanding debt?  All environments vary but the following steps set a direction in optimizing collections efforts.

Collections are tricky but if your collection strategies embed analytics, use the analytics to create proper collection treatments, and allow for ongoing test-and-learn adjustments you will optimize your collections efforts.   However, I do not want to focus on collection analytics at this time.  Why?  Because regardless of how well-oiled your collection efforts are, the single largest bad debt impact happens at the point of acquisition.

Customer Analytics in many environments provides significant competitive advantages, which culminates in minimizing waste / costs while maximizing revenue.  Many times one of the largest barriers in analytic adoption is the lack of understanding of the analytic outputs, which hinders the proper strategic application of the analytics - both short-term and long-term.

In a recently published article Online Analytics in Action I provided options for building and maintaining large online environments that allows you to analyze the date efficiently, and make the results actionable within all channels. The options presented may be extended to any large analytic data environment, and not just one that incorporates online data.

NPR recently did a story on Blue Nile, an online diamond retailer, titled Digital Bling: Diamonds For Sale Online.  Blue Nile does not have any stores, and all purchases are made online, so how do they compete with traditional diamond stores when the sensual experience of evaluating diamonds is not available online?

You may already have robust analytics that predict value, churn, bad debt & future customer interests.  You may also have a robust customer database that may be used to glean additional insights and make the analytics actionable.  That's great, but how do you determine which offer a customer should receive in all inbound and outbound channels (CSR, Web, direct mail,..)? 

I have been asked my opinion many times by clients on whether to hire strong analytic talent or spend that money on GUI-based semi-automated analytic software that the business users may use.  Both are costly so the answer should be based on ROI.

In another post 'Netflix: What were they thinking?' I summarized how Netflix used online data, a prize and the analytic community to increase revenue. 

In this blog I would like to focus on the biggest barrier in turning online behaviors into actionable and optimized customer interactions throughout all channels.  Let's start by breaking down the components of online analytics:

See a previous post for a summary of the competition - 'Netflix: What were they thinking?

It seems that after 3 years a team has won the prize.  The winning solution was submitted on June 26^th, and still must be verified, but all indications are that this solution has broken through a barrier many thought unattainable.  The winning solution predicts customers' movie interests 10%+ better than Netflix's internal system.  That does not sounds like much (10%) but given that Netflix had already squeezed about as much information out of their data as possible, and additional data sources could not be used in the competition, this was a colossal task.   

In October 2006, Netflix announced that they will pay anyone $1Million if they can improve the accuracy of their customers' suggested interest rankings (http://www.netflixprize.com/). Why? Netflix realized that if they could improve their accuracy by 10% it would be worth well over $1Million annually to the company through improving communication relevance through well targeted online movie suggestions.