Why SAP Bought Business Objects
Can a German, Franco, American alliance take on Oracle? Following SAP's announced intention to buy Business Objects for $6.79 billion, the business of business intelligence software is due for a big change. Among those changes are SAP's being forced to forego its "let's build it, rather than buy it" strategy, an acknowledgment that Oracle's $3.3 billion decision to buy Hyperion last spring looks like a great deal, a new round of betting on whether Cognos will be purchased and, of more than passing interest, almost no attention being paid to the other three letter company in business intelligence: SAP.
Business intelligence software has gone through a dramatic transition in the last five years. Five years ago, the biggest difficulty in selling the software was that few people even those in the business could explain what it was, why customers needed it and how it would benefit the end user. Now, customers are desperate for business intelligence even it they still stumble over the definition. I'd define it as software which sits on top of a company's database and helps managers analyze and sort data to bolster business strategies.
Once, this business of data analysis belonged almost solely to SAS Institute and its founder James Goodnight has been guiding the company since its beginnings in 1976. By most measures, SAS is the largest privately held software company and with $1.9 billion in 2006 revenues was still ahead of Business Objects 2006 revenues of $1.25 billion. The issues that SAS executives will dispute and are always ready to argue are that the company is a data analysis rather than business intelligence software producer, that the company's mainframe legacy translates into software which is the province of large enterprises willing to hire specially trained programmers and steep license fees assures a continuing revenue stream but blocks the path for new accounts.
The big three companies that established the business intelligence, rather than data analysis, marketplace were Hyperion, Business Objects and Cognos. The model was query tools that even a non-techie business manager could manipulate and in this era of compliance law, the systems could produce the reports CFOs and regulators craved. If there was a model, it was Google's simple query interface behind which resides a software and hardware complexity where no one wants to tread.
Hyperion's pitch was business performance management which could run independent of proprietary databases and transaction systems. The company was the first, in my opinion anyways, to talk about using the systems to look forward and develop projections rather than provide a hugely detailed business rear view mirror. The company fit well with Oracle's strategic objective to buy up the enterprise application business space a company and a set of customers at a time. The company was founded in 1981 as IMRS.
Business Objects was founded in France in 1990, has about 42,000 customers and was largely built through a long series of acquisitions with Crystal Decisions providing a key entrance into the decision reporting segment. The company struggled mightily to get all those acquisitions working together but has made significant progress in defining and creating a business dashboard where a manager can analyze the health of the company's financial operations through a few simple screens of information.
Cognos, founded in 1969 and based in Canada, did just under $1 billion ($979 million) in 2007. The company pitches itself of having the grandest vision of all in being able to combine business intelligence and performance management into one offering. Following the SAP bid for Business Objects, Cognos finds itself as the most frequently mentioned candidate for acquisition.
A couple of points. SAP has been traditionally a company far more comfortable with building its own products rather than integrating acquisitions. The integration of Business Objects will be especially difficult as for business intelligence to work, the product needs to touch nearly all areas of a customer's operations. SAP could not stand by and watch Oracle buy up the customer base for enterprise applications.
It is also difficult to think that SAS can continue to watch itself be pigeonholed into a company that has a great product if you are a bank analyzing data transactions or an airline trying to fill seats most effectively, but simply too complex for mere mortals to access and query.
And in many ways the acquisitions of Hyperion and Business Objects were a nod to the past and their big customer bases. The business intelligence big three (Hyperion, Business Objects and Cognos) or big four including SAS if I don't want one more phone call complaining I left the biggest guy out of the party were all created pre-opensource and in many cases pre-web.
A new range of companies in business intelligence designed around the web and open source is now emerging. Keep your eye on companies that are now coming out with open source-based BI, many of them using the BIRT (business intelligence and reporting tools) from the Eclipse Foundation. Pentaho, Actuate (and maybe including Squidoo) may not now be widely known vendors in the business intelligence business segment, but they were designed without the baggage of past legacy applications.