How Nasa gets the right stuff out of software

The key to success when it comes to business intelligence (BI) is keeping focus on the users of BI systems, according to a Nasa project manager.

Earlier this week, the Information Builders Summit, a user conference in Orlando, Florida was told by one speaker that the secret to successful BI projects is all about being "savvy sales people and top-notch marketers".

But according to Nasa project manager, Ronald Phelps, the key is to consult with the users and get all the definitions and processes clear from the start.

He told silicon.com: "You have to keep your community involved from day one because if you don't they change what they [want] and you'll walk in with something you think you've developed for them and it doesn't really solve the problem for them."

Making sure you are talking to the right people is also vital - and these people are often not the mangers but the people actually out on the shop floor.

Phelps said: "You need to make sure you get down to the right level. It's not about selling IT, it is showing them that IT can be one of the tools that can make their lives easier."

People in charge of these projects need to come up with a series of business definitions so that the processes can be fixed in the minds of the users. Defining the roles of each type of user also helps, he said.

Phelps said he is currently looking at better ways of delivering BI reports to Nasa staff using different types of devices.

He said: "If you look at a PC or PDA or tablet, all the display characteristics are different so one of the things we are looking at now is a capability that when someone sends a request to the server it understands what kind of unit the request is coming from and reformats [the return message] so that it shows up better."

But even with all this high-tech information gathering, Phelps is still nostalgic for the old days: "If you go back to the Apollo programme we used typewriters and a phone and our life was so much more simple.

"Today the IT people think you can analyse all this data but in many cases the data you want to get to, you don't need. People don't take the time to identify that this is the stuff that they really need to analyse and collect."