Does your planning approach to BI take more integration and customisation than pure data analysis?
Probably the simplest thing for me to do is querying and reporting against a data warehouse. That one is the least likely to require services. The state of the technology today is such that you just drop it in, do a day's course on how to use and then you start creating reports. As you go up the chain and start getting into heavier analytics such as profitability analysis you may want to think about bringing in some services as you are looking for those best practices.
Do you see services being a bigger part of your businesses going forward?
I don't see us growing services dramatically. The reason is that any successful software company that requires services needs a very healthy services partner because you can grow your software sales very fast but to get more feet on the street to do consulting requires you do the hiring, training — companies that have tried to do it all themselves have really stunted their growth.
Gartner claims that despite BI companies trying to position themselves as providing the whole gamut of services — no one provider does. In fact, every BI does something badly. What is the one thing that each of the top three BI players does badly in your opinion?
You have the ERP vendors saying we do business intelligence now — their "We do everything" claim is even bigger than people just in our space claiming they do everything. Gartner has very clearly called that into question.
There is also the question of whether you are talking about Business Intelligence or business performance management, which we do. If I was critiquing Business Objects I would say that there is an obvious omission of planning in their portfolio if you are talking about the kind of business performance management that we do.
If you were to take a look at Hyperion and say of all the things in this space that people might expect, what doesn't Hyperion do, the biggest piece would be advanced statistical analysis, something that a SAS is particularly well known for. Now do we envisage adding advanced analytics at the size we are today? Absolutely not, but in ten years when we have grown to the size we want to be then would that be something we would add? Probably.
Is business performance management a term that you would like to see usurp BI?
Yeah, we would like to see that. According to IDC figures we are number one in financial business performance management and we are number four in business intelligence. From our perspective we can say that we think it is all business performance management but if the analysts are going to count it that way we have to go out there and compete and move up from four to a better number.
Do you think this confusion over naming and definition has contributed to the relative obscurity of BI compared to other technologies?
At one level it is pretty easy to define. If you say I am in the business performance management space and it as about everything you do to drive performance of your enterprise, then a CEO is going to understand that pitch in the time it takes to go up an escalator. If I say BI to that CEO I have to start explaining all sorts of examples of how you use BI.
Going back to my example, if I walked into an organisation in 1980, I would find IBM in the finance department, DEC in the marketing department and a few ICLs and Wangs thrown in for good measure. The reason for that was that department heads were making IT purchasing decisions. You sit back now and look at that from today's perspective and think "The CFO was having hardware salesmen come into his office and making decisions between IBM or ICL — how ludicrous is that?" Eventually companies said this is ridiculous, look at all the hardware providers we have, it's a mess. They very quickly standardised and Sun, HP and Dell did very well and people like Wang and ICL didn't.
The same thing has been happening in BI for about 30 years — there are some companies out there with 23 BI technologies as a result because there were 23 different decision makers bringing that in. It has in some respects being flying underneath the radar. But now people are saying we want to start down the path of reducing the number of BI vendors we have.
Have you done any research on how many other BI vendors your customers typically have?
Well, you have heard the high of 23 from Gartner but on average we think it's around four to five.
On-demand has cut swathes through the CRM market — specifically with Salesforce.com — is there potential for a similar effect in the BI space?
I think the interest will be there but there are some technological challenges for all software companies, it is not just unique to our space. What Salesforce.com has been able to do, and it's the real secret of getting that model to work, is to create a model of running a sales force that will work across many, many companies. As they sign-up each new customer, they are not having to spend a lot to get that customer up and running. That is a bigger challenge for business performance management and business intelligence because although there are similarities in each company, each company's needs are different, and different enough that it has not been clear how to do that from a hosted perspective.
Talkback
This post has been removed by a moderator.