You may recall my blog on SAP’s farcical claims about its software’s impact on company profitability. It looks like someone with more time on their hands than me actually checked up on the figures and found these lacking, in addition to the lack of logic in the original claim. Nucleus Research, who are noted for their rigor with numbers, found that in fact that SAP customers (identified by being listed on SAP’s web site) were 20% less profitable than their peers, rather than 32% more profitable. Of course this is not quite the same thing, but it is amusing: it suggests that only SAP’s identified reference customers are relatively unprofitable. Perhaps the ones who keep quiet are doing OK? As I noted earlier, the SAP claim was deliberately skewed to exclude all financial institutions (which share the twin characteristics of being highly profitable and rarely using SAP) while anyhow the notion that the choice of your ERP systems provider is a cause of either good or bad profits is both logically flawed and also deeply amusing to those of us who have watched companies spend billions implementing SAP to little obvious effect in terms of hard business benefits.

Good on Nucleus for poking further holes in this especially egregious piece of over-marketing. Bruce Brien, CEO of Stratascope, the company that did the market research for SAP, reacted by sayng:“They’re making an implication that my numbers can’t prove, but it’s a marketing message. Companies do that all the time,” he says. Oh well, that’s all right then.

Cognos announced its full year results, notably seeing a recovery in license revenues to USD 118M in their fourth quarter (i.e. Q1 2006) after the disappointing Q4 2005 results. It was also important to note that the company closed 18 deals over a million dollars in size, which was another marked improvement on the previous quarter. Profit margins were a healthy 18%. Still, license revenue was actually down compared to the same quarter a year ago (USD 130M) while overall revenues at USD 253M for the quarter was slightly down on the same period last year. Actually shrinking is not generally a cause for celebration in a software company, so it is a measure of just how bad Cognos’ previous quarter was that these results were generally greeted with relief.

This (relative) recovery all bodes well for the broader sector, and indicates that Cognos’ stumble at the end of 2005 was to do more with company-specific issues (limited deployment of its new product line) than with any general slow-down in the business intelligence market (which just about every analyst predicts will grow at a healthy clip in 2006). In the medium term, Cognos faces the same issues as other BI suppliers: the relative saturation of the market, and the ever-growing threat from Microsoft.

Back last year I wrote about the creeping progress of Microsoft into the business intelligence arena. In CBR Madan Sheina, (one of the smartest analysts in the industry by the way), examines the latest move in this direction, the SQL Server 2005 suite’s enhanced business intelligence offerings. The new ETL offering SSIS (previously DTS) will be of interest, although its SQL Server ties may limit the take-up of this relative to database-neutral offerings. However the new Analysis Services and Reporting Services promise to ratchet up the pressure on the pure-play BI players, Business Objects, Cognos and the rest. I have long argued that the most ubiquitous BI tool is actually Excel, and that given that people already know this, an ideal BI tool for many users would be one which magically got the data they wanted out of a data warehouse directly into an Excel pivot table. Yes, there will always be a subset of power users for who this is not enough, but in the vast majority of cases this will actually do the trick. Other tools (visualization, data mining etc) would be relegated to niches if this were to happen significant niches perhaps, but niches nonetheless.

Business Objects has done well because of its semantic layer, the “universe”, which overlays something closer to a business view on top of data marts and warehouses; this imposes some maintenance overhead but this is acceptable to users since it represents the data in a more business-like form. However Business Objects has always struggled with its OLAP capability relative to competitors. Cognos by contrast, had the best OLAP tool out there in Powerplay, but a rather ordinary reporting offering. These two vendors pretty much carved up the market between them, though in a growing market there was enough room for other tools like Microstrategy, Actuate etc as well. Microsoft’s new suite poses a potent threat to most of these BI vendors, since most users do not use more than a tiny fraction of the features of a BI tool, so adding more features just to stay ahead of Microsoft is ineffective; the end users simply don’t need more features. With its low price point and “good enough” features, the Microsoft tools are likely to gradually eat into the market share of the independent vendors. Nothing dramatic will happen overnight, and the curious restraint of Microsoft from serious marketing of its tools to the enterprise will also slow progress. What was the last time you saw a webinar or advert for Analysis Services? Compare and contrast with Business Objects, which is a marketing machine.

However, just like a pack of hunting dogs wearing down a large prey animal, the Microsoft tools can just edge up on the BI vendors in reach with each release, secure in their Office base that they control what users really want: Excel.