IT Week: What impact has the increase in computing power had on the BI market?
John Kopcke: It really is enabling [wider deployments of BI]. When you look at the growth of 64 bit technology the software that stands to benefit the most is performance management and BI. I like an analogy HP uses: that if a 32 bit processor can process a city block a 64 bit processor can process the surface of the earth. That is the magnitude of computing difference and we're in a great space to exploit that.
Some commentators have talked about grid computing also enabling more in-depth analytical BI?
Grid is another option, but I think 64 bit will slow down adoption of grid as it has make it somewhat unnecessary.
What other innovations are encouraging enterprise wide deployments of reporting tools?
Role based dashboards that only present information the user needs to perform their role are a key development. For example, at Hyperion our account receivable clerks have a dashboard that is specifically geared to their role. In contrast, if you give people BI reporting tools and say "go ask questions [about your performance]", you end up with very low usage rates.
How else do you see the technology evolving?
We are working to bring more intelligence into the reporting process, and are heading towards guided analytics where we [use the data gained through BI tools] guide users through a process so they don't make mistakes and are more productive. If you look at many call centre systems now, operatives are guided through what to offer based on who the caller is and what their circumstances are. That's what guided analytics can offer at a very practical level.
Where else can this functionality be used?
It is ideally suited for supporting decisions concerning resource allocatio n. For example, Hyperion provides an application for the Miami Dolphins that assesses ticket demand and uses the information to shift tickets into the most appropriate sales channels.
Will we see BI software offered through a hosted model?
Software as a service is a technology we're watching very closely and we think it has become a good vehicle for delivering to the midmarket. We're actively investigating it and trying to understand the business model and technology implications. It is worth noting that as intriguing as the concept is the number of successful hosted companies remains very small. It's an interesting concept, but there is work still to be done.
Where does the work need to be done?
BI has always had the challenge of being unique to each customer. Lots of business processes are 80 percent the same and only 20 percent different from company to company, but analytics differs dramatically from one customer to another. Obviously to make money from on demand the cost for each incremental customer has to be substantially smaller than the revenue. If you have to customise the software for each unique customer you won't get economies of scale.
Can you get round that?
The concept of mass customisation could solve the problem. Traditionally, if you wanted to deploy role based dashboards you may need 2,000 different dashboards which would take forever to put together. But the way we do it now is to have a library of dashboard components and you pull those components together to create the dashboard. We could extend that model to an on demand basis so we provide the components to the customer and allow them to assemble the dashboard so they get the uniqueness required and we keep our costs down.
Do you see any reason why hosted might fail?
The biggest potential pitfall is that firms could be unwilling to put their most critical data outside the firewall. But that is starting to change in acceptability and the levels of security being delivered [by hosted providers] are very high.
What changes do you expect to see in the BI vendor landscape?
We've seen a lot of consolidation already in the software market and that will continue. The big question is in which area will consolidation take place? At the moment IBM, Oracle, SAP and Microsoft are in a pitch battle to be the enterprise platform. SAP with Netweaver, Oracle with Fusion, IBM with WebSphere and Microsoft with dotnet are all battling it out to be the top infrastructure player. That will be where you see the consolidation as they bring data integration, data access, workflow and all those activities into their offering. I don't think it is a logical conclusion to have a BI or BPM [business performance management] player in that mix. Most organisations are a long way off having a homogenous infrastructure so they need BI tools that will cut across all those domains. One of our big customers is 50 percent SAP and 50 percent Oracle/PeopleSoft and they aren't going to standardise any time soon. They need BPM that cuts across those two environments.
What else do you expect to see over the next five years?
Analysis of structured and unstructured data has been a topic of conversation for years, but there are signs that as xml has made unstructured information, such as documents, more structured we are getting close to providing it.
What practical benefits would the ability to analyse both types of
data have?
If you look at sales force automation systems they feature both structured data
detailing what is being sold, the size of deal, and the probability of it
closing, as well as unstructured data like call information on how meetings went
and action points for the account. The typical sales manager dashboard only
gives insight into the structured information. As a result the department has t
o go through meetings to get an understanding of the unstructured data that
really reveals the status of the account. What we'd really like is a dashboard
that automatically analyses unstructured notes as well and gives you complete
insight over the status of the account.
About John Kopcke
John Kopcke is chief technology officer (CTO) at business intelligence software vendor Hyperion.
Prior to Hyperion, he held CTO positions at data management specialist and BMW subsidiary Softlab and performance management vendor Pilot Software
