Vendor lock-in, Part II

Avoidance of vendor lock-in continues to drive much of IT spending. However, it’s not necessarily clear that buying from a number of vendors achieves the desired result of lowering costs.

Last time, I tried to make a case that a site spending $1 million on IT could get better discounts if the spend were channeled to a single vendor rather than being divided equally among 10 vendors. This is, I think, a valid point, as it generates immediate and up-front savings, but it is also probably the least of the benefits to be derived from consolidating the number of vendors at a site.

Huh?

Well, think about it. Each vendor a company engages with involves a need to manage support contracts, legal contracts, purchasing, etc. How many hours does it cost an enterprise to manage vendor contracts over the course of a year? 50 hours per vendor? 100 hours? More? If a large enterprise site with 20 major vendors must invest 100 hours per vendor to manage the relationships (it may well be twice this, by the way), what is the cost to them for doing so?

A lot.

Not yet convinced? Then let’s look at it from a different angle.

Buying all your equipment from a single vendor means you have the opportunity to buy pretested systems. How much time do you spend at your site testing systems before allowing them to be fully operational? Preconfigured systems should be pretested systems, which should have a reasonable likelihood of hitting the floor running. The bottom line: if you run a large site – or even a small one – what is the value to you if you can get your IT team out of the system integration business?

I understand that preconfigured systems can be clunkers too, but when you are a million-dollar-a-year customer, you can bet your supplier will be much more responsive than when you were a $100,000 customer.

Of course, only very large vendors can provide a full menu of products for an enterprise site, and there are only a few who can offer storage, servers, services and all the rest. The rest of them will hate this idea.

Obviously, vendor management will play an important role in making sure that something like this has a chance of really working. If you ever consider vendor consolidation as part of your IT strategy, make sure to build that strategy on two important pieces of technology.

First, make sure your prime vendor toes the line when it comes to standards adherence. Be inflexible on this: if you need to escape, the odds are much better that the switching costs (in this case, the “switching back costs”), will be much lower. Allow no proprietary data storage formats or proprietary interfaces. Second, use virtualization technologies to add an abstraction layer between users and hardware devices. The contents of a virtualized storage pool might be based on a mix of Linux-based, Windows-based and Unix-based devices from any number of vendors so, even if you begin buying from a single source, you have an escape path if the strategy doesn’t pan out.

Virtualizing storage and servers while also making standards adherence a prerequisite to any large-scale vendor relationship should go a long way toward reducing vendor lock-in fears, reducing them, in effect to the same level of concern that managers at multivendor sites have regarding portability. It may well turn out that, with appropriate vendor management and virtualized environments, such issues are no different in a consolidated environment than they are in a multivendor situation.