IT Week: Kalido has been quite critical of the current approach to data warehouse software. What is the problem with the traditional model?

Bill Hewitt: All companies now have transactional systems – be they point-of-sale systems , HR systems or procurement systems – they can all process and record business transactions. Firms are good at getting data out of those systems and into data warehouses and then serving it up to the right manager using business intelligence software. But the flaw is that once the user sees that data they go back to the IT department and say I need to reclassify this type of customer, or this product. The IT guy then has to go and write more code to rebuild the data warehouse and create new reports – typically taking about 12 weeks – by which time all the requirements have changed again.

How big a problem does this pose?

The first time a customer builds the data warehouse everything is great. But it’s not long before you want to move a product into a new product line, for example, and it has to be changed. We know of one customer that spent $1m deploying a data warehouse and when they had to change it the process cost another $150,000, and that would be the case every time they wanted to change it. Most businesses will change their strategy in a way that will affect data organisation and classification at least once a year.

Can Kalido get round this?

Our Active Information Management (AIM) approach helps overcome the problem by incorporating master data management and business modelling functionality within the data warehouse. It means you can apply the business model you are using to the data and then as the business model changes, the way the data is organised is automatically changed. For example, if next year your company decides to cut its geographical divisions from four to three by redrawing boundaries you can apply that to the data and get a different view immediately.

What advantages will the AIM approach give to companies?

It frees up a lot of IT resources as it means the business managers can easily input the business model, and the IT department doesn’t have to do the manual coding and rebuilding of the data warehouse that they had to do in the past. It also provides executives with a better way of managing the business. Business executives don’t think in terms of processes and departments, but every time they make a change to those areas it requires a change to the data structure. This means they can make changes and the data structure is changed automatically.

Is there opposition to this model among firms that only rarely change their data structure?

I’ve spoken to clients who have argued that they’ve just reorganised and got their data structure set so why do they need AIM? But what happens when you go out and buy your rival next month, or reporting requirements increase? That is when you are going to need the ability to actively manage your data structures.

About Bill Hewitt

Bill Hewitt is president and chief executive of Kalido.

He was previously president for Asia Pacific at Novell.

He has held senior positions at Peoplesoft, Brigade, Hyperion Software and ProBusiness.