topleft
topright
Enter the Member Network Zone View the Top 10 Points Leaderboard View Members Who Are Currently Online View Latest Member Activity

Featured Members


Member Network Zone

Expert Blog Comments

Google App Store Open for Business
I will be keeping an eye on this, it does not seem like Google fails at much but I have yet to be ov...
State Updates Extortion Laws to Cover Identifiable Info
Hopefully, common sense changes like these will become part of every states laws. Unfortunately, the...
HP Tries On a New Brand Image: ‘Let’s Do Amazing’
I like it. It's a good counter to IBM's Smarter Planet campaign, which really captured the idea that...
The Trouble with Big IT
Michael - In reading this I kept going back to an IT project risk model that I know of as coming vi...
Agility? Surely You Jest...
Ellen, Esther, Peter, Tom, Thank you for your comments. I've continued the discussion in my next ...
The Digital Chauffeur Print E-mail
Article Index
The Digital Chauffeur
Be Agile, and Be Willing to Accept Risk
Data Explosion with Dynamic Pricing

Data Explosion with Dynamic Pricing


The use of dynamic information for pricing will result in a massive data storage requirement.


Currently, the typical auto policy record contains around 500 bytes of data to describe the insured plus a similar amount for each additional driver. Each covered vehicle requires about 350 bytes. For the majority of policies, no changes are made during a 6-month term and no claims are filed, so only a single record is stored. This record stores such information as the rating classification code, policy term, coverages and amounts, and other basic policy information. However, for a dynamic pricing model to succeed, records will need to be stored that identify location information that changes throughout the day. While this basic data will require only such things as a timestamp, duration, and some form of identification key, the possibility of producing hundreds of records in a single day per insured exposure exists.


An executive from one carrier interviewed said that the carrier captures GPS data once per second while the vehicle is in motion. This has produced a massive database of structured data. Such an approach will result in considerably greater data volumes than most carriers are accustomed to managing as well as significant analytical challenges and expense increases. Carriers will need to factor the cost of additional storage as well as transmission costs into their cost/benefit analyses. Furthermore, larger volumes of data take longer to analyze; therefore, insurers must invest in faster analytic engines that leverage the parallel processing capabilities offered by modern databases and computing platforms.


Decisions will need to be made about whether to store event-driven data or interval-based data. Some preprocessing of data should be done within the insured vehicle prior to a data transfer, but any summarization will result in the loss of detail that may provide a source of competitive advantage.




Comment on this article
RSS comments

Only registered users can write comments.
Please login or register.



 
< Previous   Next >




White Paper Library