Re: How to determine database market share?

"Paul" <paul_at_see.my.sig.com> wrote in message news:qjpgl1davpp9vfra8i8omsjfs57qi4bfcj_at_4ax.com...
>
>
>
> Hi all,
>

<snip for brevity>

Hmmm.... Do you realize that there is a strong correlation between the decline of pirates in the Caribbean and global warming. As the number of Caribbean Pirates has declined since the 1700's the global temperature has risen. Do the research its true.

Its funny, customers are like penguins. They sit on the shore or ice burg until one of their kind jumps into the water. Then they all jump in at once.

So your theory that job adverts for a particular RDBMS technology has a correlation to "market share" (whatever that is) may well be true. Now should one adjust for median income differences? What about foreign currency conversions with median incomes between countries? Often the range of size of the RDBMS (volume or transaction rate or both) are mentioned in the advert. Should we normalize for the TPS volume or gross bit size? What about ease of maintenance? If Oracle makes it easier to maintain its Database does that mean it is losing market share? If I can manage 10 X more Oracle stuff with the same amount of database administrators is Oracle losing "market share". Why would they commit market suicide like that? Wouldn't it be better if they made the database harder to administer so that they could gain market share? (requiring more DBA's) In fact the best strategy would be to make it so administration is not scalable but make it super easy to develop applications for it. Then we would drive DBA adverts and developer adverts way up and hence market share.

Let's ignore the term "market share" it is a pretty meaningless term. I mean what is "market share" do we count a db on a PDA the same as on a mainframe even though they are each one instance? If someone downloads a free copy of Oracle for development (which one can do) does that count towards market share? It doesn't generate any revenue and actually costs Oracle something. On the other hand, it helps Oracle gain customer loyalty and potentially lock in someone who has spent time learning the product. (which one could argue is gaining "market share") If I create a dozen instances on a 1 CPU machine how does that affect market share? What about a 128 node RAC system? Is it the total size on disk? Is it how many CPUs? Is it how much was paid for it?

You need to define what problem you are trying to solve. You need to define your terms more precisely. You need to indicate what value the information you seek would have. If it is risk avoidance then the company with the most or the least "market share" should be avoided. The smallest "market share" has the most risk that it will go away. The largest "market share" has the least incentive to make it easier for you or solving your problems if they arise because you are small potatoes to them. Go for the one in the middle, it will probably survive, handle your problems with mediocrity, and not go anywhere in the gray haze of the middle.

Personally, "market share" is one small factor in a buying decision.

I am sure you will respond with your unprofessional profanity because I've pointed out some things to consider would it have been "better" if we all had just praised you and exhausted your erudite musings? ( I suspect I'll get the ol' FU and then he goes into the Kill File, to be ignored for eternity.)

Jim Received on Sat Oct 22 2005 - 01:31:18 CDT