保险行业BI案例概览-CSDN博客

You are a business person. You may also be a technologist. Regardless of which role or roles you fill, either you or your boss wants hard cost and ROI data on any investment the company is going to make, especially in regards to software. You need to understand how much money a particular software investment is going to save you, how your implementation is going to accelerate, how quickly it will provide a business solution to the business users, among others.. While this is understandable and absolutely vital with any investment, you need to look at business intelligence (BI) initiatives from this ROI angle, but also a very different, futuristic angle.

At this point, you may be hesitant to hear advice from a vendor advising not to examine ROI during a software purchase. That's not exactly what I'm advising. There are so many more things to consider when going through a BI/Data Warehouse purchasing process. In my humble opinion, there, are more important variables than "cold, hard implementation numbers", which often are not revealed for two to five years after implementation. As with any investment, those "cold, hard numbers" depend heavily on your company's willingness to change. This willingness to adapt is, without question, the most vital thing your company must do in order to make a BI/Data Warehouse initiative successful from an ROI standpoint.

This is extremely true for the insurance industry; an industry known more for its plodding pace than for agility. Don't get me wrong, you can absolutely measure ROI for things like:

· How long will it take to get in production?
· How much money will I save by going with a pre-built model over building one internally?
· Will I be able to maintain the new solution with fewer resources?
· How much will it cost me to customize the models that are in the market today?
· What does one solution save me over another?
· What's the cost of developing everything in house?
· How much maintenance cost will I save over my current solution?

A vendor can and should be willing to help you with the questions above and questions similar to those, but the hard ROI questions must be internalized and thought through. Often these questions are technology and carrier independent. Internal carrier questions may refer to:

· What decisions have we made based on insufficient data? What was the cost or profit of those decisions?
· How many decisions were delayed due to lack of information?
· What's the opportunity cost of not having quick access to information at every level of the organization?
· How fast can I go to market with a new product now? What are the bottle necks in developing new products? Where could I use better/faster information? How will this speed my go to market?
· Do I know if I'm profitable at a policy level? Or do I know if I'm profitable at a policy group level (Male drivers, age 21 that drive Pintos).
· Do I know what policies to continue or stop writing? How much money am I losing from these bad policies? How much money am I gaining from good policies?
· Do you know and understand your renewal rates? By Geography, LOB, Agent/Channel, etc...
· Are my agents driving high profit business or high loss business - which agents, geographies, etc?

Do you know the answer to the above questions? Often just thinking about those questions brings on night terrors and cold, sweaty palms from line of business managers, underwriters and IT directors. Primarily because of three reasons: You can't get to the information quickly to make those decisions, you've made decisions in past based on partial data, and not least important is that your company simply doesn't ask these questions. McKinsey & Company alluded to this in their Managing IT in a Downturn report in 2008 by relaying, "Few companies have successfully capitalized on the explosion in data in recent years. Often this information, residing in separate IT systems or spread across different business units, has never been mined for insights that could add value."[1] This is especially true in insurers that have multiple admin systems (policy and claims) and multiple disparate analytic solutions within departments like actuary and claims. Compounding this issue for insurers is that they are one of the world's biggest hoarders of data and typically only actuarial departments take advantage of the data.

Kknowing all this, how can you position your company to take advantage of a BI initiative? You must first be able to answer the company-specific questions above to justify a BI initiative.I don't use the term "initiative" lightly. Buying and using BI to make your business more profitable takes initiative. It takes executive initiative and it takes IT initiative, claims initiative and underwriting initiative, and actuarial initiative... Every group must be involved in the decision to make use of BI. This doesn't mean that every group needs to be involved in examining the technology, (except for vertical solution sets), but they should have some internal benchmark questions answered for their department that can give IT a leg up as they begin the arduous determination process.

Therefore, you must understand if your company has the initiative to make BI adoption necessary. This involves having commitment and understanding from each of your departments. Each department also must develop their own metrics in which faster access to information, and the visual analysis of data, can help them answer. Set a standard with department-specific benchmarks to encourage participation in this exercise. With that said, the following should be considered and answered affirmatively before making BI/Data warehouse decisions:

· Does your management team (executive, line managers and mid managers) have the initiative, wherewithal, and the desire to change the way and the speed in which they make decisions today?
· What are the questions that each group does not have answers for today? Either no answers whatsoever or you may have received the answers, but they were so late in coming to you that it didn't matter when you received them. Make departments accountable. Ask them to reach outside their normal day-to-day activities and ask questions that they don't know the answer to today, but could have they had access to specific data.
· Have you empowered your mid-level management (claims management, underwriting management) teams to make decisions when new paradigms present themselves?

You should ask these questions under any BI initiative and you should provide your people with the tools to answer the questions that are most nagging.

Lastly, while attaching a purchase ROI to justify BI initiatives is not the single, recommended approach,, I would like to state for the record that I'm a big fan of ROI and payback analysis - I wouldn't be a supporter analytics if I were not. So, by all means, create an implementation ROI model and understand the cost and time associated with implementing, customizing and maintaining of your BI and data warehouse system. Just be sure to wrap that ROI model with the real business questions of why you are moving forward with a BI initiative.

[1] McKinsey & Company, "Managing IT in a Downturn: Beyond Cost Cutting", Fall 2008