Saturday, August 4, 2012

Is Business Intelligence dead?

Neil Raden posted recently an article entitled " BI is dead! long live BI".  This adds to the trend of declaring some technology or trend as dead:    Earlier this year I wrote a post in this blog, entitled: "Is mapReduce dying?" in which I also indicated some references to claims such as  "SOA is dead" or "ERP is dead". Actually nothing in the IT industry dies quite easily,   I guess that COBOL, which is about my age, is still alive  today, and may outlive me.   What Neil claims is that the success of BI was much less than the hype created around it, and its adoption was around 10-20% in large organization depends on the survey.   This is consistent with other opinions,  James Standen reported  on a survey that found that the the actual adoption is lot less than the percentage claimed  by BI vendors are. The picture at the top of this page is borrowed from his post "Business Intelligence adoption low and falling".     
Neil claims that BI does have a future, but its future is in doing it as part of a decision management framework, not as a stand-along service consumed by people.   When talking about decision management, it has a role to analyze historical data, but doing it within a context of a specific decision.  This can be used to provide input to decision or to reinforce or refute assumptions.  

I think that this is an interesting perspective and in a way has some similarity to event processing which also transitions from stand alone technology to part of a bigger game.   I have discussed it in the past, and still stand by the conclusion that stand alone event processing will stay as a niche, but the mainstream will be part of bigger games.  I have investigated recently the current list of "bigger games"and will write on it in one of the future posts,  however, I see the same observation as valid for BI,  current BI will remain a niche, while the mainstream will move towards embedded BI.   More -later. 

Sunday, July 29, 2012

Event processing in practice


My old friend Roy Schulte from Gartner, who has been the first analyst that published analyst reports about event processing, and one of the founding fathers of the event processing community and EPTS has published a "personal blog" on the complexevents site with the provocative title:  " does anybody care about event processing?".   Roy expresses his opinion (that I've heard from him several times before) about the event processing concept and market. Roy's short answer to the question he asks is YES, but many of those who are doing event processing are not aware that they are doing event processing.   He mentions two main issues:
  • Terminology:  the term event is manifested in many different ways, while the community centered around the current products have narrow interpretation
  • Market:  Most of the event processing is custom code, or built into vertical products, and not the general purpose event processing platforms, thus he had to scale down his prediction for revenue that count the event processing platform market.
Roy suggests that EPTS will decouple itself from the focus on the product oriented functionality and concentrate on event processing as technology in the broad sense.   

I concur with Roy's observations (we have discussed it last year when I visited him in the Gartner office);  I have written before about the build-buy trade-off,  based on chapter 1 in the event processing manifesto, a chapter that Roy was one of its authors.      I also noted in some of the feedback to my book that I got from developers who told me that the book helped them a lot in architectural thinking and clarity of concepts, but they also said that they'll continue to develop event processing applications in custom code.  

Roy's Blog has been written in the context of the discussions on EPTS and online magazine,  and I have expressed my opinion in the same spirit as Roy --  the online magazine (and EPTS) should appeal to anybody that are doing anything with any kind of event in any form.  This should capture all the "event processing in practice" in the universe.  More news about the online magazine -- soon (the editorial board is now being created, which contains both old members of the community and new blood of practitioners).