By: Bill Baker, Microsoft General Manager and Distinguished Engineer

I had the opportunity to speak on a panel at the most recent TDWI World Conference. The discussion topic was "The Next Big Thing in BI." As expected, more than a few 'science projects' came up. I'm all for science projects. I'm still itching to get working on that holographic data slicer/dicer. But I think there is a more pressing need in most of the companies that I have the privilege to meet. I think the next big thing in BI is Return on Investment.

Five years ago, many, if not most, of my conversations with customers revolved around 'Why BI?' Now, not so much. The more common conversation is about driving BI deeper and wider into organizations. This, by-the-way, has always been our vision for BI. Customers ask us how to get from 200 BI users to 2,000 users or from 2,000 to 20,000. I believe this comes from two angles. First, companies and organizations have had good practical experiences with BI in recent years. They truly believe in the power of Business Intelligence to improve their companies. Second, they are sensitive to the ROI of BI projects.

Companies implementing BI systems and applications face fixed costs and per-user, marginal, costs. Regardless of the number of users, a BI project will involve extensive up-front work with the business community, design, implementation, hardware, software and services. The costs of these items vary relatively little with the size of the user population. A reasonably priced four-way server can handle hundreds of users. But its cost is far less than the cost of the up-front collaboration with the business users or the implementation costs of designing and building schemas, acquiring and cleaning data or any of the other non-hardware, non-software expenses. Information from AMR indicates that hardware spend in BI systems is around fifteen percent. Yes, doubling the user population will drive up server costs, but these still tend to pale in comparison to the other up-front expenses.

Category spending in BI/PM 2007 Billions Percent Internal labor and HC 7.46 31.4% Software 5.87 24.7% Integration 3.86 16.2% Hardware 3.67 15.4% Outsourced services 2.92 12.3% 23.78 Source: AMR Research 2007

I've met CIOs who have spent millions on a data warehouse only to realize 200 users. Even at one million in up-front expenses, the per-user cost of the back-end is five-thousand dollars if you only have two-hundred users. That only drives down to two-thousand dollars per user at five-hundred users. At that volume the marginal costs per user for client software or licenses dominate the fixed costs. The key point here is that some companies are frustrated by the relatively low usage of the BI/DW facility they have built. They aren't achieving the widespread benefits that BI promises, and they have very high per-user costs.

My thesis is that driving up the number of users increases return. This is core to our BI vision; more people making more immediate, better informed decisions leads to better companies and organizations. In other words, return is partially a function of usage. It's true that bad usage of data creates risk and high-volume bad usage is even worse. There is an old joke about the pilot who announces to the passengers that they are lost, but making good time. There is no substitute for getting the design right, getting the data right, training your users and in general providing them the least amount of data and the most amount of guidance. In the future, I'll cover our thoughts about empowering employees while maintaining alignment to strategy and corporate standards.

In future articles I will delve into strategies that companies can employ to drive up BI usage. Items I will cover include:

• The role of familiar tools, like Excel and SharePoint, to accelerate adoption
• The use of scorecards as the natural entry point to BI for many users
• The role of self-service BI to increase relevance as well as capture creativity
• Using data lineage to build trust in data
• Embedded BI to put the right amount of data and tools where the users already work
• Spanning the empowered-aligned spectrum
• Change management and other social tools
• Techniques to drive down costs, the other part of the ROI equation

I'm particularly interested in the social barriers to BI adoption. Please feel free to share your thoughts on that or any of these items with me at billbak@microsoft.com. Side note: I will be on sabbatical in April and May, but will be responding to email in May.