Organizations with enterprise resource planning (ERP) systems face a common question when implementing or upgrading business intelligence (BI) programs: Choose the existing ERP vendor's BI system -- often touted as a more straightforward implementation? Or look at third-party BI software, which may promise different or more specialized functionality?
The bad news is that there's no simple answer. The good news is that for organizations with encompassing investments in ERP, the BI options are better than ever. Most ERP systems now come with a variety of built-in, integrated and available BI solutions, and there are still reputable pure-play, third-party BI vendors finding success. There's been a lot of action over the last several years.
"Many of the largest standalone vendors have been acquired. Oracle kicked it off by buying Hyperion, SAP acquired Business Objects, and IBM followed with Cognos -- these moves attempted to consolidate BI within the ERP world," explained R. "Ray" Wang, a partner in enterprise strategy with Altimeter Group.
Despite all of this consolidation, however, there are still strong independent BI vendors out there, he added. And all the recent interest in BI has also meant that even the smallest ERP vendors have focused on expanding their BI capabilities. That adds up to lots of choices. The big question, though, is how do you make the choice between what your ERP vendor offers and third-party BI software?
ERP and BI software experts say, "Follow the data"
"Clients who tend to have homogeneous environments with their ERP tend to go with an ERP provider's solution; however, that's only if they are comparable in capability," Wang said. A more capable third-party solution can get the nod if the ERP vendors' solution is too limiting.
But what if there's a lot of data coming from mixed sources? Many ERP-focused companies have business processes that utilize non-ERP data. What then? In that case, Wang noted, "we see a lot more use of specialist solutions when the data is heterogeneous and when customers have advanced use cases that go beyond simple reporting."
Boris Evelson, principal analyst of BI with Cambridge, Mass.-based Forrester Research, voiced a similar sentiment. He's hearing more and more requests from companies that have a strong ERP presence. In many of those cases, he said, there is little reason to look elsewhere for BI -- unless you have very specific requirements. And many companies do.
"What if most of your reports require operations that SQL cannot do?" Evelson said. "Then you have to look at tools that have their own programming languages -- that are more powerful and flexible than SQL. Another example is in-memory analytics, which allow end users to explore without the limitations of a data model."
Indeed, in-memory analytics pose some interesting new opportunities. Where traditional RDBMS and OLAP architectures require the "pre-discovery" of data -- through data modeling, integration or warehouses -- in-memory models can cross-reference attributes, creating indexes in the form of a virtual data model on the fly, Evelson noted. Still, the data has to be reasonably clean: It's not a panacea for data quality initiatives. And the downside for ERP-focused companies is that the end results can be even less predictable than traditional BI.
ERP and BI software lines increasingly blurry
Making the choice also requires looking ahead for both long-term requirements and system innovation. Jeff Woods, managing vice president of ERP and SCM for Stamford, Conn.-based Gartner Inc., said that within a few years, the bigger trend will be components of ERP applications that you won't be able to purchase without buying the analytics environment as well -- the analytic environment will be integral to the execution of the applications.
"The BI components might remain standalone, but they'll become required runtime components of the ERP applications," Woods said.
So does this mean that integrated ERP BI has the advantage over third-party options?
"It depends on the degree to which the ERP vendor reengineers the ERP suite to take advantage of analytics. If they deliver presentation-level embedded analytics, then this is interesting, but not that interesting," Woods explained. "If they deliver a reengineering of business processes because of the presence of analytics functionality, then we're talking about something that is transformative, and it would be very difficult for someone to capture without using the embedded functionality of the ERP vendor."
The point? If employees can use embedded analytics within the course of their decision making process, they are more likely to put the company's data to work. If a question requiring human intervention arises, the human might not be able to track down the answer in a separate BI system, but if an employee can extract the answer through the ERP application, so much the better -- choices can made quickly, backed up by evidence delivered by data. Experts also say embedded analytics can help ignite the imaginations of business executives, which in turn helps make the business case for BI.
While embedded analytics are only currently available in bits and pieces, there's been a gradual inclusion of BI in ERP. "You increasingly see more and more components of their products relying on a BI point of view," said John Hagerty, vice president and research fellow of BI and EPM for Boston-based AMR Research. "They are delivering a lot more dashboards, more reporting, more analysis as part of their applications."
Still, ERP-based BI is not a slam-dunk proposition.
"The reason is that ERP isn't the be-all and end--all; it's that it's often part of an overall system -- companies have other business applications they built themselves or bought from other vendors, and they recognize that the data from all of these applications potentially has to be integrated together to make analysis more complete," Hagerty explained.
"At this point, folks are recognizing they have to step up and make BI investments beyond what ERP provides," he added. "ERP is important data, but it's not the only type of data in a business."
There's another interesting situation created by market consolidation, he said. A lot of BI solutions that ERP vendors have acquired were previously standalone solutions. The BI-ERP integration came later. Can't these previously standalone tools still connect to almost any form of data? To find the real answer, Hagerty recommends, look to how it's marketed.
"If a BI tool is sold built-in, it may be more limited," he said. "If it's marketed as an open-ended tool [also sold separately], it can probably look at most any data store."