A world of trade-offs

Vendors and architects need to better understand the business case that drives design decisions.

In my conversations with Net technology vendors, I am often amused at the absolute terms they use. They talk about “24x7 uptime” and “massive scalability” as though they were attributes that you could just buy off the shelf. And, of course, they are sure that the shelf you can buy them off of is their own.

They also tend to assume that total scalability and reliability are so essential to e-business that cost/benefit analyses can be trivialized. Wrong. They can’t.

There are two main reasons that close scrutiny of business case is essential to smart technology acquisition and deployment. The first reason is that these things all cost money. Technology vendors rarely want to discuss this well-kept secret. It’s amazing how many new-product announcements I am bombarded with that don’t include retail pricing or licensing. The cost of this revolutionary new solution, I am told, is “to be determined.” Well, then, fellas, my assessment of your product’s value proposition is “to be determined” too, okey doke?

Vendors try to sneak around this issue by claiming that their technology is so indispensable that the denominator in any cost/benefit calculation clearly dwarfs the unknown numerator. I guess that means that intellectual-property attorneys should feel free to charge these same vendors several million dollars to file patents for them. After all, patents are “indispensable” to the high-tech business, aren’t they?

No, the fact is that every business only has a finite amount of money to spend on technology and—now more than ever—it must choose how to spend those finite dollars very wisely. Sometimes I think Wall Street’s overvaluation of tech stocks has gone to these people’s heads. It’s as if they’ve been seduced into believing that just because investors are willing to pour seemingly unlimited capital into technology stocks that there is somehow a corollary that corporations are willing to pour unlimited capital into technology purchases.

That’s far from the truth. Tech budgets at mainstream corporations are not climbing significantly. Payroll, marketing costs, and legal fees (can you say “regulatory compliance?”) are actually growing faster than IT budgets at many organizations. I’m not an economist by profession, but I would venture to say that the run-up in tech investments has more to do with money needing a home than it does with the inherent value of dotcom business models. Just look at where they’re investing their newfound millions. (Hint: it begins with “bricks” and ends with “mortar.”) So, technology vendors shouldn’t let their market caps go to their heads when they have to sell stuff to stingy IT purchasers.

A MATTER OF CHOICE

Okay, here is the second reason that business case needs to get more attention from sellers—and buyers—of technology. In IT, as in life, it often turns out that two goals that are individually quite desirable are actually mutually incompatible. Becoming a rock star or a doctor. Getting your degree from Harvard or starting the world’s biggest software company. Controlling the worldwide browser market or steering clear of the Justice Department. Marrying Uma Thurman or Heather Locklear. Being right or being President.

Scalability and reliability present IT architects with a similar choice. To increase your Web site’s scalability, for example, you throw a load balancer between your router and a whole bunch of servers. Congratulations, you’ve just introduced a single point of failure.

In fact, achieving high reliability typically requires having at least two of everything. That can be a very expensive proposition—especially if you’re trying to build a very high-capacity infrastructure at the same time. You often have to choose between having two of something to double your scale or two of something to provide hot failover.

How do you choose? Easy. You find out which is more important to your company’s top and bottom lines.

Here’s a perfect example. WebSideStory tracks site traffic for its customers. Its business model requires it to be able to track any amount of traffic for any number of customers. Its site design is all about scale. If it can’t capture your site’s stats for a few minutes or an hour, it’s not really a big deal. You’ll still have the other 23 hours of stats to go by. It’s not that WebSide wants to build an unreliable site. On the contrary, it spends a lot of time and money on maximizing its reliability. But, when push comes to shove, it can tolerate a tracking outage here and there.

That’s obviously a different scenario than a financial-services site. If I can’t buy IBM when it’s at an 18-month low, I will be very unhappy. You better be super reliable, or I’ll take my business elsewhere.

My advice to buyers and sellers in today’s action-packed technology market: stop talking in absolutes. There is no free lunch, and there are no panaceas. There are only value propositions that vary in how compelling they are. That’s what the business of IT is really about.