Two interesting statements from a recent report in American Banker (password required):

• In the decade from 1996 through 2005, banks added an average of 1,484 branches a year
• There is a direct correlation between the location of a branch and its success or failure in amassing deposits

Going by this, retail banks could certainly use the optimization methodology, we used recently (see related post) to map branch location to ‘desirable customer segments, to increase average deposit per branch

There is a very insightful article in Business Intelligence Review which states that almost all the BI tools available in the market are not based on any sound scientific theory of ‘decision making’

Articles references Herbert Simon’s decision making theory:

“It is work of choosing issues that require attention, setting goals, finding or designing suitable courses of action, and evaluating and choosing among alternative actions. The first three of these activities - fixing agendas, setting goals and designing actions - are usually called problem solving; the last, evaluating and choosing, is usually called decision-making.”

and criticizes the fact that the popular BI tools do not incorporate any kind of intelligence when it comes to problem solving methodologies and wrongly assume that providing a lot of manipulated and packaged data for reviewing will lead to better decisions. A case is made for decision-model driven analytics software to replace current the data-driven packages.

Is it always a good idea to push more services to customers?

Our analysis shows the answer to be a very big ‘No’.

Recent analysis of mobile data services at a major US wireless operator proved this clearly. For the high- risk credit classes, the high data charges from mobile content drove the users to stop paying their bills, leading to a rise in involuntary churn of customers.

As the total data charges in the first two bills increase, there is a huge jump in the involuntary churn rate especially after $15 mark. Almost half of the customers with $30 and up in their first two bills are charging-off.

(click on the thumbnail to view graph)

So what’s the solution? In our opinion it is two-fold:

1. Rigorous analytics to identify the risky segments
2. Proactive customer retention strategies like providing SMS/email alerts to inform customers about excessive usage

Not only will this strategy decrease the involuntary churn rates, it will also act as a customer satisfaction tool and help customers make more informed decisions.

One would think that Hollywood is an unlikely application area for predictive analytics. Not quite, according to some recent research.

Research by Wharton professors shows how predictive modeling can be used in Hollywood, to better evaluate the scripts and predict financial success of the movie.

The core idea is to make the green-lighting process, where professional readers cull movie scripts for production, more objective and rules based, using a technique called natural language processing.

A great use case of analytics helping create an objective decision making framework in arguably one of the most subjective domains.

You can get the entire paper from here

As jurisdictions across the US adopt electronic voting, the government administrations, vendors and users are climbing their respective learning curves in implementing new processes, developing new solutions and adopting new protocols at the polling locations respectively. As is almost always the case with large scale deployment of new technology solutions, there have been issues (see earlier related post).

Cook County in Illinois had its share of problems (go here for more details), where we recently had the opportunity to work with a team on determining the root causes for delays in transmission of election results. Given the significant implications of our findings to the players involved (the Cook County Board of Elections, the technology vendor), it was critical that the analysis be entirely data driven. To us the experience was very much like being part of a “Whodunit?” investigation.

We adopted a classic hypothesis driven approach to divide and conquer the problem with hypotheses ranging the gamut of possibilities across technology driven causes, process failures and governance gaps and break-downs.

Network applications and systems maintain rich data and that was the bedrock of bulk of our analysis. We could conduct rigorous and rich analysis of every component in the system that kept logs. The systems we dealt with, ranged from small embedded client devices, switches in the wireless network, servers in the central office to databases where results were tallied. The data sources were diverse, but the beauty was the utter simplicity of the analysis driven almost entirely by the clean structure adopted for the analysis - problems always seem simple to solve if divided into logical and small chunks.

The summary of our analysis was a time series plot of how cartridges moved through the system. This was a powerful visual, as it was entirely data driven. Yet, it made obvious the evolution of scenarios that caused a lot of stress on election night.

From a visualization perspective, we utilized a tool that enabled us to study correlations between the location of a voting device and the odds of successful transmission. We used a Microsoft tool called MapPoint and were thoroughly impressed. More details about it in another posting.

Of late, a lot of our clients are having conversations with us about using predictive analytics (here is a good definition), in an effort to create sustainable information advantage (see Diamond’s whitepaper on the topic).

A recent report by TDWI on best practices in predictive analytics, based on survey responses from 833 industry practitioners, provides some interesting data points:

1. 64% of respondents are either exploring or developing predictive analytics programs
2. Among those who have implemented predictive analytics programs, two thirds (66%) say that it has provided ‘high’ business value
3. Cross-sell/upsell, campaign management, customer acquisition, budgeting and forecasting and attrition/churn are the top 5 application areas for predictive analytics
4. Median investment of companies with successful predictive analytics programs is $1M annually (60% of which is on resources, 20% on software, 15 % on hardware)
5. 56% of a predictive analytics project time is spent on definition, data exploration and preparation
6. Complexity of models, poor data-quality, high processing expense, lack of expertise, lack of interoperability of systems and pricing of software and hardware were the main barriers cited for organizations to venture into predictive analytics
7. Report recommends hiring business savvy modelers, and creating specialized ‘Analytics Datamarts’ from the data warehouse where modelers can develop models

You can get the full report from here

Malcolm Galdwell (of Tipping Point, and Blink fame) uses the definition put forward by national-security expert Gregory Treverton to highlight the difference between a puzzle and a mystery, in a recent New Yorker article:

“Osama bin Laden’s whereabouts are a puzzle. We can’t find him because we don’t have enough information. The key to the puzzle will probably come from someone close to bin Laden, and until we can find that source bin Laden will remain at large.

The problem of what would happen in Iraq after the toppling of Saddam Hussein was, by contrast, a mystery. It wasn’t a question that had a simple, factual answer. Mysteries require judgments and the assessment of uncertainty, and the hard part is not that we have too little information but that we have too much.”

The core thesis of the article is that we take too many problems to be puzzles (and try to gather more data), whereas in most cases we will get to the answers if we think of them as mysteries (and do more analysis of the available data). Avinash and folks from Juice analytics tend to agree that the problem with many practitioners of web analytics and customer analysis is a ‘puzzle’ attitude which results in producing lots of reports and metrics, many of which do not provide any actionable insights. For businesses that have spent millions of dollars on data warehouses and ERP systems over the last decade, the problem clearly falls into the ‘mystery’ domain.

However, the question that will become more important over next couple of years is:what should be the organization’s strategy to harness actionable insights from all the available data?

An automated data-mining approach (a technology heavy solution which BI vendors will push and claim can be performed with little understanding of the techniques) or a more traditional hypothesis-driven approach (which relies on domain expertise as much as technology)?

PS: It is instructive to understand both sides of the argument (here and here) from a similar debate which has been going on in the field of genomics and proteomics, who have amassed large datasets of their own, in last few years.

How can my company develop sustainable advantage and market leadership by leveraging information and analytics?

For executives who are faced with this challenge, Diamond has published a whitepaper based on our work at various clients across industries. The whitepaper, uses some interesting examples from Allstate, Catalina Marketing, Kraft, The Hartford among others to illustrate the 5 key drivers, which Diamond has identified, that companies need to manage, to build and sustain an information advantage:

1. Picking critical areas to compete using an information and analytics strategy based on evolving industry and value chain dynamics.
2. Cost effectively and systematically applying data-driven analytics and fact-based decision-making to improve business performance.
3. Accelerating fact-based, data-driven decision making cycles across all levels of the business to improve market reaction times.
4. Exploiting technology advancements and using new delivery processes to alter the economics of collecting, storing, managing, accessing, and analyzing data.
5. Re-thinking how to organize business and technology resources to get the most insight from information, at the appropriate speed, for the lowest cost.

You can download the whole paper here. You can also visit John Sviokla’s (co-author of this whitepaper) insightful and thought-provoking blog.