Vendor lock-in, Part I

In days of old, when knights were bold, and interoperability was not invented…no one worried about vendor lock-in because, after all, that was all there was.

Software running on your DEC machine was pretty much useless if your company decided to shift over to a Data General environment, so few people did. Applications for an IBM mainframe didn’t necessarily work on other lines of IBM machines, and certainly didn’t run on any other company’s boxes (and even if they did, not many of us understood that EBCDIC stuff – and how to translate it into ASCII – anyway). So what you had pretty much determined what you would have. In most cases, very little switching went on.

Now, of course, the whole idea of vendor lock-in is very much on people’s minds, although I am beginning to wonder if that concern is still legitimate or is just the result of a series of misperceptions.

Vendor lock-in assumes that once a set of products is installed, the incompatibilities associated with proprietary environments make the switching costs too severe to be economically viable. To avoid that, sites have moved to open environments, systems that, even if they are not completely “open,” at least can claim a degree of openness because they subscribe to a set of open standards. This typically means they can then buy servers from more than one supplier, storage from more than one supplier, and so forth.

Managers sometimes call this “splitting wallet share” or something similar, and think it keeps them from being under some vendor’s thumb. At the very least, in theory, they have the option of playing one vendor off against another to get a better price.

But maybe not. It may well be that IT sites are just as locked-in if they run a multivendor operation as they would be if they operated a single-vendor shop, and that they are somewhat the worse-off because of it.

Consolidating around a single vendor carries with it certain advantages, one of which, curiously enough, may be in pricing. On the surface this sounds counterintuitive – after all how can we get better prices if only one vendor bids?

Here’s how. Assume you spend $1 million each year on IT purchases, and that you split the buying equally among 10 vendors. Each vendor then loves you to the tune of $100,000 per year, which should be the basis for a fine relationship. Now assume an alternative approach, one where you consolidate all spending with a single vendor. Might not that vendor’s heartfelt appreciation for you be expressed in a better pricing structure now that you have embarked on this more significant level of commitment?

The answer, quite clearly is “yes,” but whether that answer is loud enough to drown out the lock-in issue isn’t clear. Still, saving an added 5% to 15% on your capital expenses isn’t a bad deal if your IT budget is tight.

Look deeper into the concept and it turns out that CAPEX isn’t the only place where advantages lie. There are OPEX advantages as well, and they seem to pop up in the most unexpected places. More on that next time.