If you want proof that business intelligence still has not made it everywhere yet, take a look at what typically happens when a company is purchased by a private equity (PE) firm, whose goal is typically to turn the ship around to (more) profitability and growth and sell in 3-5 years.  Executives are incented with company performance and often put their own skin (i.e., money) in the game.  This creates the effect of a sprint out of what was a marathon.  Waste is removed from the company equation, and then some. 

So, what happens to business intelligence initiatives when this growing movement of private equity ownership happens to a company?  From what I gather, 50% of the time, BI is either substantially reduced or eliminated.  Ouch.  Do the new owners think the company can achieve its new goals better without BI?  Do they see BI as only relevant for companies with goals beyond 5 years?  Some clearly do.

We are now several years into the emergence of the PE trend and are seeing some of the aftereffects of the PE strategies.  One of them is the RE-introduction of BI.  In these cases, either BI was (1) providing value, but it was imperceptible by the PE firm or (2) not providing good value.

Either way, in my experience, companies never ultimately do not do business intelligence.  It may be called something else.  It may be dormant for a while.  It may be floundering and in need of revitalization.  PE firms take note - let's evaluate the BI in place before making changes.  BI can, and ultimately does, work with a strategy of streamlined, 3-5 year horizons.

Technorati tags: data warehouse, Business Intelligence, private equity

The biggest business intelligence trend in 2009 will be a flight to perceived value. This will include the return of ROI measurements for what we do. This has happened before and will be nothing new to industry veterans who remember 2001 and 2002, when good economic times burst with the dot-com bubble and world events.

Almost instantly vaporized were the speculative, technology-first business intelligence projects, replaced with a return to the cold, hard questions of "Why are we doing this? No, really - why?". The largely silent response quelled business intelligence progression for a while, but with some confidence returning in 2003, some of the handcuffs were loosened. However, it has never returned to pre-2001 glory, and that is a good thing.

Anytime a discipline gets too self-absorbed and fails to remember its long-term existence depends on contribution to the bottom line of the business, it is set up to fail. However, along with the good of the refocusing of business intelligence come the challenges. Then and now, business intelligence, when not inextricably tied to its business results, loses its way within an organization. The value of the underlying infrastructure, providing well-performing high-quality data to applications, processes and people tends to get overlooked when BI is positioned as technology. BI's role in the value chain can be lost.

Ultimately, BI is more important than it has been for quite some time. The techniques that are going to emerge in 2009 are those that play to perceived value and the easiest form of that is reducing the expense of BI. These include SaaS BI, open source and quite possibly data warehouse appliances (for replacement deals.) These will eventually include the cloud .
Whatever the method, BI must provide perceived and recognized value.

Fortunately, I was exposed to the ROI requirement in my first business intelligence project back in 1994, thought it to be a great idea even when not required, and include it routinely now.

Technorati tags: data, Business Intelligence, return on investment, ROI

I thought I would start off the new year with somewhat of a recurring theme today - saving money. And specifically, one approach to information management that is being touted as very aligned with this strategy is open source. Although open source is generally working its way into organizations from other angles like JBoss and Linux, open source DBMS is a second-tier port of open source into many organizations.

"Enterprise DB is to Postgres what Red Hat is to Linux" is how the folks at Enterprise DB put it. They augment and expand the open source DBMS Postgres with Oracle, and other, functionality. They use open source Postgres and keep to its updates, unlike other vendors like Netezza and Greenplum, who have picked up a version of Postgres at some point and forked off. I'm not necessarily putting either approach forward as a virtue, just as the facts. Other open source DBMS models include the more popular MySQL, which is owned by Sun, who could close-source it at some point if desired. Some of the newer features in MySQL (MySQL Enterprise) are only available to paying customers.

Note some of the legacy of Postgres in the picture. "Post"gres stands for post-Ingress.

Sometimes the functionality in Enterprise DB trails Oracle by a year depending on the need in the marketplace. However, it is a tight fit and code port effort has to do with the depth of recent Oracle SQL extensions used. Enterprise DB is targeting the SMB market, who may not need all the bells and whistles of closed source DBMS. They offer support of open source Postgres and 2 levels of support (code is free) for Enterprise DB, their variation of Postgres. At $4500 for an annual subscription, it's not free, but enticing.

I will be responding to an ask-the-expert question here on the B-Eye-Network very soon on whether open source can be used to run an enterprise data warehouse. Look for it in the video blogs here.
The open source DBMS are mostly utilized in OLTP. However, I think that is more a result of more DBMS in OLTP. So can open source work in EDW? Yes, it can and yes it is. I've been ramping up my team's capabilities around open source. I'll have some cautions about open source EDW when I get to the video.

Technorati tags: data, Business Intelligence, open source, Enterprise DB,MySQL