Focus on Analytics and BI: Understanding The Interplay Between SAP, Microsoft, SAS, IBM, and Oracle

There’s no more an important and convoluted marketplace than business intelligence and analytics, a hodge-podge of tools, solutions, applications, technology, and other “stuff” that often defies analyst taxonomies and baffles customers.

This places an additional burden on trying to sort through the recent coverage about the SAS Institute duking it out with IBM, a recent deal between SAP and Microsoft in the enterprise performance management space, and what all this means for customers and competitors in this crowded, confusing marketplace.

What isn’t confusing is the recognition that a new era of analysis is dawning on the enterprise, one in which the proliferation of data and data types (relational, non-structured, image, etc.), data sources (inter-enterprise, intra-enterprise, consumers, smart devices, etc.), and users (everyone with a brain and a task or goal) is making it painfully obvious that the old problem of turning data into information has become orders of magnitude more difficult even as it is becoming similarly more essential.

Into the widening breach between data and action have ridden more companies than you probably would ever want to consider in an RFP: in addition to the big names above, my friends at IDC name another 45 companies that pull in 80 percent of the revenues for analytics alone. The 50th company is that ubiquitous “other”, commanding 20 percent of the market and a very respectable 7.8 percent compound growth rate from 2006-2008.

And that’s just the analytics side of the business. Borrowing IDC’s taxonomy, this BI/analytics market space actually encompasses almost a dozen distinct and not-so-distinct categories: financial performance management, supply chain analytics, CRM analytics, production planning, services operations, and workforce analytics on the applications side. Then there are the BI tools (three sub-categories) and of course the business warehouse side (with two sub-categories, all stained with gallons of red ink).

In short, what IDC calls Business Analytics looks at everything from end-user tools to finished, packaged applications to underlying ETL and database technology. It’s a pretty messy market: for the most part this tortured analytical framework is the result of a market taxonomy that is more about what vendors have to sell customers than what customers need to get the job done.

What the customer – as in end-user/business analyst – wants is more and more found in the performance management sector of the market, and therein lies perhaps the most exciting market competition in enterprise software (actually, the database side is pretty exciting too, but in the interest of focus, we’ll leave that aside for now.) And this is where the significance of the recent SAP and Microsoft deal comes into play.

Microsoft has been great at the tools side of its business, and SQL Server is getting kudos across the industry for its scalability, embedded analytics, and generally excellent design. Sharepoint is also surging as a platform for analytics and collaboration. But Microsoft has lagged in its ability to create the high-end performance management apps – planning, consolidation, budgeting, etc. – that require a deep understanding of the inner workings of a business.

(In all fairness, there’s a decent amount of business process knowledge in Microsoft’s Dynamics applications business, but it’s hampered by three problems: a Microsoft policy that refuses to favor Dynamics over other partners in key parts of the business, a lack of understanding among members of the Microsoft BI team that they could and should tap Dynamics’ expertise despite the ban on favoring Dynamics, and the fact that much of this business process knowledge comes from the partner community, and is not part of the core of the Dynamics offering.)

That business process knowledge is something that SAP has been first cultivating, then acquiring, during the years that Microsoft has been building is technology platform. In fact, the recent deal between the two centers around a former Microsoft partner, OutlookSoft, that SAP acquired in 2007 in order to compete with Oracle’s acquisition of Hyperion earlier that year. Which adds some color to the reasoning behind why Microsoft likes this particular part of the SAP portfolio: it’s built on the Microsoft stack, and therefore pulls a decent amount of Microsoft technology into the enterprise, at the expense of a common enemy of SAP and Microsoft: Oracle.

This performance management deal is further evidence of a widening partnership between Redmond and Waldorf (and Palo Alto) that is refining the demarcation lines in the overall market. SAP and Microsoft have a growing list of common enemies, and a growing set of reasons to find common ground in the applications space. SAS is one of those enemies, and to judge by the coverage SAS is getting in what’s left of the mainstream press, one would think SAS is the company to beat in the market. But is it?

BusinessWeek most recently anointed SAS as a major contender against IBM, and by extension SAP and Oracle, claiming that SAS’s positive license revenue growth was in contrast to SAP and Oracle’s negatives. But comparing SAS’ revenues to Oracle and SAP’s overall revenues is like comparing apples to a fruit salad: both SAP and Oracle have much broader software portfolios than SAS, and in fact the majority of both companies’ revenues are not in the same competitive arena as SAS at all.

If you look specifically at the business analytics market, where SAP, Oracle, and SAS compete, SAP and Oracle outpaced SAS’s growth in the 2006-2008 timeframe that IDC’s most recent report covers. As did IBM. And Microsoft. Of course, SAP and IBM had recent mega-acquisitions, which the silver-tongued CEO of SAS, James Goodnight, poo-pooed in his BusinessWeek 1500 words of fame. Despite Goodnight’s dismissal of these acquisitions, the fact is that BOBJ and Cognos didn’t actually compete directly with SAS, and the IDC numbers for 2006-2008, which include BOBJ in SAP’s 2008 data, show that SAP, and possibly Oracle, was outpacing SAS’s revenue growth without having to depend on acquisitions to do so.

Meanwhile, just to make sure SAS keeps its eyes open, IBM bought traditional rival SPSS last summer, and is hiring business consultants and domain experts to fill out its bench and go directly after SAS’s tool+consultant business model.  It’s not easy being SAS these days.

But, to close the loop here, what SAS has to offer is important to the market, and its birthright as the consummate data analysts’ tool is unparalleled. (I was both a SAS and an SPSS programmer back 20+ years ago when my business card said “statistical programmer”, and SAS was the consummate data analysis toolset.) What SAS can offer the analyst who knows what he or she needs to analyze is pretty much top drawer in the analytics market. As long as you know what you need to know about your data, SAS can help.

This capability, however, turns out to be more of a liability than an asset as the proliferation of data and users of data grows across the enterprise. The problem is that SAS, like most of the BI/analytics firms that its rivals have acquired, is hampered by the fact that the bulk of SAS’ business process knowledge is not embedded in the software itself: in the SAS market, business process knowledge is most often baked in by SAS’ army of consultants, who adapt its more standard tools to the requirements of the individual company. This and their licensing structure makes the resulting solution relatively expensive to own and maintain, especially when considering more packaged solutions.

This is why IBM+Cognos and SAP+BOBJ were powerful combinations: two companies with deep business process knowledge (IBM’s via its Global Services arm, SAP via its Business Suite). And this is why Goodnight is wrong about the impact of these acquisitions. And it is also why Microsoft is turning to SAP to provide the business-aware solution for enterprise performance management that Microsoft failed to provide in its now defunct PerformancePoint product.

In an era when a one terabyte hard drive costs less than $100, the issue of the proliferation of data, data types, and data analysts is only going to be more profound. The moves of the likes of IBM, Oracle, SAP, Microsoft, and SAS are sometimes obscured by the confusing terminology the surrounds the different components of the solutions to this issue.

But, whether obscured by nomenclature or not, the essence of all these recent activities is the same: the biggest battleground in the enterprise is shaping up to be about converting a growing mountain of data into information, and there is a common realization that this requires better proscriptive solutions, a la the SAP/Microsoft deal, that are pre-packaged and relatively easy to consume. How each of these vendors responds to this challenge will determine who wins and who loses the most important battle for customer mindshare and walletshare in recent memory.



4 thoughts on “Focus on Analytics and BI: Understanding The Interplay Between SAP, Microsoft, SAS, IBM, and Oracle

  1. Pingback: Tweets that mention Focus on Analytics and BI: Understanding The Interplay Between SAP, Microsoft, SAS, IBM, and Oracle « Enterprise Matters -- Topsy.com

  2. Josh

    Great article as always.

    Another aspect of the Microsoft business model that hampers its growth in business applications as a whole, not just analytics, is that it views “content” to be the domain of its partners. As a result it never acquires the in-depth knowledge of business processes. As you point out, the exception is the Dynamics team.

    What I find to be the biggest drawback of nearly all analytics tools at the moment is that they are either backward-looking – analyze history – or very thin forward-looking tools with almost no consequence evaluation, for example demand forecasting. Even in demand forecasting there is little to connect a financial forecast to a unit forecast, let alone generating a statement of supply from the demand.

    Of course in some areas there are attempts to “connect the dots” using correlation between variables, but this isn’t causality. Even worse, it often results in the assumption at the high level that correlation is the same is causality. Very little is done, for example, to test the feasibility of the assumptions made in generating an annual operating plan.

    I think this is where the breakthrough value of analytics will be realised.

  3. Pingback: Not Your Omi’s SAP « Enterprise Matters

  4. Pingback: Not Your Omi’s SAP

Leave a Reply

Your email address will not be published. Required fields are marked *