Continuity in BI Solutions – the User Perspective Revisited

Ever so often in BI development one must go a fair distance from A to B and has to draw up a roadmap for doing so. That plan should be agreed upon between Business and IT – between users and developers. And there are several pitfalls in the process.
Here I will discuss the case of continuity as experienced from a user perspective. In some cases it is evident that continuity must be preserved when moving from the old BI solution, A, to the new BI solution, B: continuity of the collected data, historical and new, - and continuity of presentation in that calculated KPI’s remain intact with effectively identical business definitions.
In other cases disenchantment with the existing, old solution may be such that it may be tempting to opt for a “revolutionary” new solution without regard for real cross-solution continuity. From an IT perspective this may seem the more attractive and easier way to go, and Business may be persuaded to go along. Or so it may seem. According to my experience, things are rarely as simple as that.
The problems are that initially Business may not fully comprehend and understand the scope of a new BI solution, and that the Business representatives at the defining stage may not really be representative of all Business users.
As an IT person one should ask the right questions and those questions are not always just the simple and direct questions. In this case it is important to know what triggered the need to go from A to B. Are we looking at a native and immediate desire for a new BI solution, or are we really handling a spinoff project from a change of ERP system, a merger or acquisition, or a high level policy change in the company? In any of the latter cases, even if Business superficially agrees to a whole new BI solution one should be quite wary of possible complications. After all, the existing solution has been and still is supporting business needs in the short term. Any excitement about a new solution may quickly disappear when faced with the challenges of implementing it. How deeply rooted is dissatisfaction with the current solution? It may actually prove quite shallow!The solution then, I propose to be, that IT should really probe the level of user dissatisfaction with the current solution before settling for taking off with both legs to leap into the future and construct an entirely new BI solution to replace the existing one. And unless completely satisfied that the old solution will be missed be no-one, the prudent thing to do is to dissect the old solution and to document any changes, especially where features may be discontinued. It may seem tedious at times, but it actually has a number of advantages for IT too. Any such documentation and understanding is useful as part of the specification of the new solution. It also helps to clearly define and emphasize what is actually new features in the solution to be. And eventually, any difference in data presentation between the old and the new solution is just a potential error as long as it remains unexplained, - once understood and explained it becomes an improvement and an extra feature! - Last, but not least, a thorough understanding of the old solution is an immense help when constructing and implementing the new. Contrary to the leap into the future described above, it is like a long walk where one foot always remain in contact with the ground. And after all, when literally going somewhere, I – as most people – do tend to arrive faster and in better shape after a steady walk than after a series of leaps and jumps. Agree? - Stay tuned here at morlin’s BI blog for more deliberations on both technical and organisational aspects of BI!

A Simple Business Proposition (Motor Insurance)

Less than a week ago Kim Rud-Petersen, Fair Forsikring’s [Insurance] executive of the Business division blogged [in Danish] on Borsen.dk: Kim Rud-Petersen’s subject was the extreme conservatism and very slow - even to the point of laziness - business development within the insurance industry at large. I used to work with DW/BI for an insurance company for several years, and I strongly agree with Kim Rud-Petersen. Let me give a simple example of what I mean; a case of how insurance companies might very usefully activate some of their more or less dormant data to the immediate mutual benefit of their private motoring customers - and themselves!
If I want to trade my car, papers and data are important. With a comprehensive service history and knowledge of accident repairs, trading the car is easier and the price is usually significantly higher, see e.g. www.bilpriser.dk. Of course, some information – e.g. about serious prior damages – will depress the price, but the trend is clear-cut and important: With less uncertainty, the risk premium is reduced and a better deal is possible for both an honest buyer and an honest seller. But when you’re a private trader there’s a snag. You have really no access to solid information about repaired damages, that is privileged information for the auto trade with their proprietary data bases and professional experience. [At least in Denmark; in Sweden a comprehensive and public database is maintained. In the US a less comprehensive privately organised database exists.]
Now enter the insurance industry: Suppose I wanted to sell my car and advertised it at e.g. www.bilbasen.dk. Wouldn’t it be nice to have the option of going to the website of my auto insurer and simply release for public showing the damage and repair history of my car? I could even link directly from my advertisement to my insurance company, and thus make my car more attractive with a more comprehensive description immediately available to prospective buyers. Perhaps I might even be willing to pay some small amount for the privilege of attracting viewers and generating traffic to my insurance company’s website!
But that is only one side of the coin. Try taking a look from the side of the insurance company. As it is now, the insurer hears nothing before the car is sold and the business of insuring it is discontinued. With the described feature, the insurance company gets early warning when a customer releases damage and repair history of his car for public viewing. The company is then able to proactively take steps towards protecting the business; ideally business may actually be increased – doubled - when the company is able to hold on to both the owner/seller (with a new car) and the traded car (with a new owner)!!
This was only an appetizer. Rest assured that serious money is involved overall, making it worthwhile for a progressive and competent insurance company to establish such a feature. It is not even particularly difficult to develop and calculate the idea as a business case. I once did it over 10 pages, but that is a story too long for a BI blog. I am personally convinced the idea will be realised; only question is when and by whom. It would have to be a company genuinely concerned with helping and empowering its private customers, even at the risk of annoying somewhat the auto trade by slightly infringing upon their perceived privileges. I hope some time to be able to tell that story. Stay tuned here at morlin’s BI blog for any possible developments.