Gartner: Search technology in business intelligence not the answer |
 |
By Christina Torode, Senior News Writer
18 Mar 2009 | SearchCIO-Midmarket.com |
|
|
SearchCIO-Midmarket.com recently caught up with Nigel Rayner, research vice president at Gartner Inc., during the Gartner Business Intelligence Summit in National Harbor, Md. Here, Rayner talks about the dangers of adding search technology to the business intelligence (BI) strategy mix. He also gives his take on why several technologies are converging in the BI space and lays out how to make a business case for performance management projects in tough economic times.
Some experts believe search technology is a next-generation business intelligence tool for structured and unstructured data. Where do you see search playing a role in BI strategy development?
Rayner: Well, in a lot of cases no one is using the business intelligence tools they have. The worst solution is throwing another tool at your users. The problem isn't that they don't have access to information or tools; they already have too much information, and that's just in the structured BI world. Now you want to couple it with unstructured data? That's a whole load of garbage coming from the outside world.
The answer is putting information in context to help business people make decisions, which is performance management. They need to be able to do modeling, simulation and look at past trends. That's why analytic applications are so important. They provide a framework, a methodology for making a decision. They will also give you the process support for scenarios like bringing on a new product and the related modeling, marketing and product development costs. [Analytic applications] will couple this financial information with the information on the potential market for this new product … including what people are saying in the unstructured world.
So introducing search technology into the mix will hurt a BI strategy?
Rayner: You should be using search to help you get competitive intelligence, market intelligence. But all you're doing is throwing more information at business users, and they'll be saying, "You give me 10 bits of information and tell me to make a decision? Well how do I know what I should act on unless you give me an environment in which I can model a business outcome?" That is modeling based on guesswork.
How can people use search as part of BI, then?
Rayner: Part of the problem with traditional BI is that it's very focused on structured information. Search can help with getting access to the vast amount of structured information you have. But, you can argue that if you organized your information properly, you wouldn't need [search]. Also, if you knew which operational information you cared about, which actually influences corporate performance, you would be able to reflect that in master and meta data relationships and you shouldn't need search.
There is a lot of talk about the convergence of business intelligence, business applications and business performance management. What is driving that?
Rayner: I think it's where the vendors and the technology is headed. It's not yet necessarily where the users are headed. So this is being driven by the big vendors saying, "Let's buy all of these components across the whole stack, put them together, and everyone will say, 'Let's buy that.'" It's a combination of the vendors being visionary, but also trying to lock in their customers.
Where is this convergence becoming apparent now?
Rayner: You had separate worlds of traditional business intelligence focused on the infrastructure and technology pieces like information management, data warehousing, BI platform, reporting, analytics, querying and some advanced analytics. Then there was the world of specialists that did performance management or analytic applications. Hyperion was a great example of that for finance. What they were doing was creating analytic applications to address the needs of the CFO and management team: budgeting and planning, financial consolidation and reporting. Then you had the transactional [business application] world of ERP and CRM, as well as things like business process management.
|
|
|
|
|
Search can help with getting access to the vast amount of structured information you have. But, you can argue
that if you
organized your information properly, you wouldn't need [search].
Nigel Rayner
research vice president, Gartner Inc.
|
|
|
|
|
|
|
|
|
|
All of those things are starting to come together. What's happening the most is we're seeing the emergence of these analytic applications. So what Hyperion was doing -- and a few niche vendors -- has blossomed into corporate performance management, which is focused on the need of upper management and the CFO. The corporate performance management space is a $2 billion market and is growing at a rate of about 19%.
At the same time, finance needs applications for budgeting, planning, forecasting plus analytics. The problem is customers don't get that [functionality] from their ERP systems, so that's why BI has extended out. But you can also look at it as the performance management space has sucked in BI.
Vendors like SAP and Oracle are finally building in what users thought they were going to get when they invested millions in an ERP, which is great reporting and great business management tools. But the problem is that [the ERP vendors] are building from the bottom up. They built a data model for business applications, with a layer of analytic applications on top of it.
With Oracle's acquisition of Hyperion and SAP's acquisition of BusinessObjects, they both recognized that that approach of building it from bottom up and adding it onto the ERP was not really getting onto the broader BI market.
Now [with the acquisitions] they're putting together that end-to-end vision: the idea of a range of performance management applications supported by an underlying BI platform linked into the business applications.
[ERP vendors] are also building process-driven analytics on top of transactional applications and their BI infrastructure. So in many ways, that shift in thinking on the part of SAP and Oracle have driven the market towards convergence.
But that's not necessarily what customers want?
Rayner: Customers aren't buying that. They aren't saying, "We want it all and we want it all integrated." They're still at the point of buying individual point solutions. It makes sense to have a BI platform and CPM suite tightly integrated, but the buyers are still buying BI platforms, CPM suites separately. And as new areas of analytic applications emerge like sales performance management, the head of sales will go out and buy that. That person should be working with IT and with the CFO to make sure that what they buy works with what they have, but you know they won't.
People are looking for ways to cut costs. Is there a simple way of using corporate performance management to do that?
Rayner: There are two areas they should be focused on: planning, budgeting and forecasting; and profitability monitoring and optimization. Both are very simple and can be implemented in three to six months.
A small implementation can be done for $200,000 to $500,000. CPM applications don't require a big infrastructure; the implementation is pretty straightforward, and they can hook into anything … ERP systems or data warehouses. [CPM application] investments can be justified because you're giving everyone the tools they need for better budgeting, planning and forecasting. Most people are still using spreadsheets, and that takes too long; it's too much of a manual effort. You can build a business case that will get a return in 12 to 18 months for corporate performance management.
Let us know what you think about the story; email: Christina Torode, Senior News Writer.
|
|
|
|
