A recent flurry of news articles in the database industry has reinforced my perceptions that the software titans have lost touch with current market needs.
The PR machines from IBM and Oracle have been each claiming to have the fastest database and hardware platform on the market today. Database vendors battling each other on who has the fastest or the most features is nothing new and has been going on for a couple of decades. But what has changed is “who cares?” Of course it does matter if you are a techie and you are evaluating which platform has those characteristics. It also matters to industry analyst and consultants that are evaluating and recommending the “best” database platform one can buy. But maybe the key criterion for future database market share growth is not the best but what is the “best buy.”
Just as what is the “best” is a relative term and why “best buy” is what one should look at, so too is the TCO (Total Cost of Ownership) studies that are published by the same people touting the best. Many TCO comparison studies assume a massive enterprise database deployment and if that is indeed what you are implementing they are relevant but if not “who cares?”
Key market segments for future database growth are: emerging markets, SMB (small to medium size businesses) and business groups (referred to as departmental computing in the old days) within an enterprise). These groups may need business intelligence, predictive analytics, performance management or operational reporting. They may need to get data from their enterprise applications but also structured and unstructured data from customers, suppliers or other employees that they work with.
These groups are looking for a database that meets their functional needs at a reasonable price and that does not require a fleet of gurus to design, develop, deploy and maintain. The functionality battle is not what database offers the most but rather who meets the threshold needed, i.e. good enough. And if you do not have a large enterprise deployment the “good enough” threshold means plenty good enough.
In a post I wrote in May, 2007 “Database Market Share: The Rich get Richer” made three predictions:
These predictions have held up well and still apply today.
Besides open source databases (a distribution method rather than a new type of database), columnar databases, in-memory databases and database appliances are all making in-roads to the database market because many offer great value. Although the prospect of these products displacing the titans is way in the future, if ever, but many will enjoy significant growth based on their value to their customers.
And, if they do prove to be successful, look for more titan acquisitions.