What are the top 3 things to consider when purchasing a business intelligence solution?

1. Prioritized Business Needs
This is driven by the vision of what we need the BI solution to do for the company and also provides the paramaters of why the company would invest in the BI solution. If the needs are not clearly defined the ability to know if a solution is heading in the right direction is seriously in jeopardy from the outset. An agreed Prioritization of requirements is necessary since there is usually a multitude of stakeholders who often have their own biases.

2. Fit to the Organization
The selected technology stack, (whether it is from one vendor or a blend of different software vendors or even customized software or open source) must fit within the organizations technical culture this is both in the user community that needs to use the solution and the IT community that needs to support the solution.

3. Budget
I put budget 3rd on the list in that while this may be tight constraint, often the BI solution needn’t be limited to a single investment. So while the Budget may be minimal at first, if ROI is proven early in the adoption of the solution, it can unlock additional organization investment for growing the solution. Additionally, if during the analysis of the potential solutions, it is shown that ROI is significant then additional investment above an original budget threshold could be justifiable.

Jeff, Having seen BI implementations ranging from the very good to the very bad, I think a BI solution needs to be:

1) User friendly, self-help oriented and designed for the users and not for IT experts, because constantly needing an IT person to run SQL, MDX or complex reports is a slow and unproductive process for everyone.

Having representatives of all business areas attend the vendor presentations and participate in the selection process is fundamental to develop commitment and support. Of course, IT leads and facilitates the process. In order to succeed, a BI project needs to be an equal partnership between IT and all the business sectors that will use the solution.

2) The solution needs to accommodate a flexible design of the meta-data to be "business intuitive" because you do not want business folks to rely on analysts as intermediaries to translate BI reports into useful information through extensive and time consuming manipulation of data exported to spreadsheets.

3) The BI tool needs to be fast and interactive (like OLAP cubes) to be able to drill down, drill through and filter, so everyone can uncover the root cause of problems in a very few minutes.

Finally, while cost may be important to some folks, Return on the BI Investment is the ultimate measure of success. If the selection and configuration of the BI solution allows the company to beat the competition, increase market share, revenue and profit everybody wins.

Regards, Bill

Hi Jeff,

I think that the Top 3 questions to consider when thinking about BI are:

1. What am I trying to accomplish? And what if this changes?
Business value needs to drive a business intelligence implementation. If it doesn't, it's a dead project and a waste of time. At the outset, you need to have a clear sense of what you're trying to achieve, where that data comes from, and how it will be consumed.

That said, you also want to make sure that you pick a solution that can easily adapt to change in direction, new data sources, etc. So keep in mind that you need something that will be agile and adaptable to future requirements.

2. What's my budget today and for the ongoing future?
BI can be very expensive, especially if you're going the traditional on-premise route - stitching together a lot of BI components by hand and maintaining it all over time. If your budget and IT resources are leaner, there are still high quality solutions - there are new, modern solutions like SaaS BI that provide equivalent enterprise power, but at a far lower price.

3. What can my organization handle in terms of capabilities, culture, and process?
Can you dedicate resources from IT now and for the foreseeable future? Do you have a group of analysts that can respond to business user's needs? If so, traditional solutions are just fine.
Or do you need something with fewer IT resource demands and have an organization of business users that wants to do reports itself? If so, you might want to consider alternative BI solutions.

In the interest of full disclosure, I work for a SaaS BI company, Birst (www.birst.com), so I'm inclined to it. However, there are other choices out there, too - like in-memory analytics, depending on what your requirements are.

And here's some more info on traditional BI vs. SaaS BI http://www.birst.com/bi101/traditional-bi-vs-saas.shtml