Multi-dimensional Analytics of Individuals

When social meets psychology

In 2010 Tal Yarkoni, an academic at the university of Colorado, Boulder, suggested in an 
article that it might be possible to gauge a person's personality through their writing
by tracking the 'Big Five' personality traits - openness, conscientiousness, extraversion,
agreeableness and neuroticism. Mr Yarkoni argued that extroversion correlated with "bar', while neurotics were found to use the words "awful", "lazy" and, "depressing". But other
findings were more remarkable. Trusting types were more likely to use the word "summer", while more co-operative beings favoured the word "unusual".

A group of researchers at IBM's Almaden Research Centre in San Jose, California, had picked up on Mr. Yarkoni's idea and applied it to Twitter. The team, lead by Eben Haber, 
hope to discover the "deep psychological profiles" of tweeters. Analysing three months'
worth of data from 90m users, they argue that so far they have been able to gauge 
someone's personality reasonably well from 50 tweets, and even better from 200. 

Multi-dimensional Analytics of Individuals

Why they are doing this? Well, all of these research aim to derive indivisual psycho-profile
out of social media sources. What if we are able to augment to more advanced dimensions such as human's value, needs analysis, social behavior as base foundational elements of analytical framework. Value measurements represented by individual's intrinsic value such as self-transcend, hedonism, conservation etc., and Needs might be mapped with "curiosity",
"self-express", "harmony" etc. Social behavior may reveal each individual's bevavior such
as "morning tweeter" etc. 

By doing this, current technology may represent individual psycho-profile through the lens of these four elements - personality, needs, values and social behavior. We may bring the different insights and consequent dimension to this equation but nontheless, ultimate goal would be identify each individual's unique psychological traits.

From firm's point of view, these insights could be regarded as different lever to drive different offer target to different response. Further more, if we are able to put another insight such as "network potential" into each customer data, we may offer tailored messages as timely basis (since we know the social behavior) via preferred channel from these enhanced digital profiles of individuals.

**The postings on this site are my own and don't necessarily represent IBM's positions, strategies or opinions.

Big Data Workshop for Insurance Executives

Insurers as experienced risk planners

I had an opportunity to lead the interesting idea exploration workshop with South Korea leading insurance executives on Big Data and its implication to Life Insurance Industry in September.

Session kicked off by sharing the case where South Africa’s largest short-term insurance company uses predictive analytics to uncover a major insurance fraud syndicate, saving millions on fraudulent claims and resolve legitimate claims 70 times faster than before.

Like most insurers around the world, this company was losing millions of dollars paying out fraudulent claims every year. To improve its bottom line and enhance customer satisfaction, the company needed to detect and stop insurance fraud early in the claims process.

Solution gained the ability to spot fraud early with an advanced analytics solution that detects patterns in near real-time and captures data from incoming claims, assesses each claim against identified risk factors and segments claims to five risk categories, separating higher-risk cases from low-risk claims.


Results was stunning. Identified a major fraud ring less than 30 days after implementation and saved more than $2.5M in payouts to fraudulent customers, and nearly $5M in total repudiations. Further more, reduced claims processing time on low-risk claims by nearly 90%, and resolves legitimate claims 70 times faster than before.


Effective Channel Management


Next question was, how can I make my call centre more productive, while providing better customer service?

If we are able to combines data about the individual customer with each contact center agent’s specific skills, expertise and past performance to optimize the routing of calls. Current technologies and consulting capabilities designed a “matching-engine” which leverages this combination of customer insight, agent profiles and real-time analytics to provide “individual-level” decisioning and assignment of calls not available in most contact centers applications. By doing this, this company provided more personalized customer/agent interaction so sales yields increased by 29%.

Insurance Customer Journey Map

The argument was, insurers should shift their gravity to the role as experienced risk planner who may design the customer's journey map centered around customer's life event and every action customer should focused on locking in a profitable customer for life. Journey map could be explored with customer based on customer experiences and life event and examined and explored from the context of each age period such as twenties, thirties etc... and equation between expenses and residual incomes.

So this journey map possibly illustrate the future key event such as Auto purchasing, wedding, first long-term saving discussion, retirement planning, realtor, home purchasing and wealth transfer etc... through out the customer's life event.

Cause a fundamental shift in the Insurance business model

This can be realized by leveraging Big Data to model future risk development for customers based on experience of similar people. For example, non-discrete data such as video of home, audio of agent conversation may be used to complete asset inventory and propose coverage. Probably, real time predictive operational analytics to empower better business decisions regarding customer service, marketing, and infrastructure deployment.

Or group insurers assess the risk of individual insureds based on actual behaviors and score them for rating, behavior coaching, and up-sell. And not only build capital advantage through rapid compliance by predicting/planning for future regulation changes, but also process the advanced case management for clinical decision support, underwriting, and claims investigation.

These observations lead us to imply that it will cause a fundamental shift in the Insurance business model : Predict customer life stage evolution before it happens and proactively market.

** 여기에 포스팅한 내용은 개인 차원의 것이며, IBM의 공식적인 입장, 전략, 의견을 반드시 대표하는 것은 아닙니다