As I think I’ve probably said before, I don’t get many freebies as a blogger but one that I do get and I really appreciate is my review copy of the BI Survey. It being that time of year, I got the latest edition – the BI Survey 8 – last week along with an email encouraging me to blog about it, and who am I to refuse a request to blog about something as fascinating as this? It’s a whopping 489 pages long so I can’t even begin to summarise it, but there are some points regarding Microsoft BI that I’d like to pick up on.
In general, the findings show the Microsoft BI stack as a solid and successful suite, extremely good value for money, but by no means a stellar performer. There’s a long section at the end of the report showing a wide range of technical and project-related KPI ratings (for example query performance, business benefits achieved, quality of support, cost of deployment and so on) and in almost every respect Analysis Services and Reporting Services come out in the middle of the rankings. This doesn’t really surprise me much: let’s face it, what we’ve got works well, but from a technical perspective there’s not been much new and exciting in the world of Microsoft BI for a while now (although with Gemini and Madison coming soon that will change), the platform in general still has some glaring holes and overlaps, and as the PerformancePoint debacle showed recently Microsoft’s overall BI strategy is somewhat confused. In fact, I suspect Microsoft’s entire BI strategy is not a BI strategy at all but a get-people-to-upgrade-to-the-latest-version-of-Office strategy, but I digress…
For Analysis Services in particular, the survey showed 15% of those surveyed were on AS2K, 79.6% were on AS2005 and 3.5% on AS2008. Given that most people must have been surveyed last summer, well before the RTM of AS2008 that’s pretty good and with the migration path from 2005 to 2008 very smooth I should think the AS2008 percentage will look much better next year. Analysis Services remains the top-ranked BI tool used against Microsoft’s own databases, as you’d expect since it’s essentially free with SQL Server, but it also comes in a close third for Oracle (while Oracle’s own BI tools come in 12th!), second for IBM, third for Teradata and top for Open Source databases.
In terms of the client tools used with Analysis Services, just over 70% of people were using one of Microsoft’s own tools – mostly pivot tables and Reporting Services. 25% of SSAS users had Proclarity, and interestingly there were more people using the old free Excel addin than PerformancePoint , and a sizeable minority still using Data Analyzer. There’s clearly a lot of demand for a client tool from Microsoft but from the looks of things most people are still stuck with Office 2003; this just adds weight to my argument that Microsoft coupling its BI strategy so closely to Office might help Office adoption rates but has a serious negative impact on the success of BI strategy itself… sorry… there I go again… As far as third-party tools go Panorama is still the #1 vendor, but it only has an 11% share and only 22% of SSAS users overall had any kind of third party tool in use. Incidentally Panorama was treated as a BI vendor in its own right for the purposes of the survey and came out top in a lot of the high-level KPIs including the overall KPI ranking – they obviously have some very enthusiastic customers.
So who, apart from Panorama, seems to be doing well? Qlikview certainly is, which at least validates Microsoft’s decision to go after that market with Gemini; Microstrategy does well too. Essbase comes out badly with great query performance offset by poor support and product quality; Cognos, apart from TM1, doesn’t seem to do too well either. Clearly it’s the vendors who are 100% focused on BI that are the most successful, which is as you’d expect.