Ease of use and the exasperation factor

Despite all the praise vendors heap on themselves about ease of use and total cost of ownership, it still remains caveat emptor.

Every product, according to its makers, is easy to use. And, for the folks who designed it, I’m sure it is.

But, say what they will, so many products that should be easy to use, just aren’t. This “exasperation factor” – the sigh you heave when you hit that brick wall – undoubtedly causes many would-be buyers of a demo package just to leave, never to return. Still, vendors seem to be blissfully unaware of the opportunities that they are losing.

My recent experience with a claimed “enterprise-class” file replication package (which shall remain nameless) is characteristic of so many offerings today.

The package runs on Windows and provides, it says, “powerful, intelligent” features for mission-critical applications allowing replication across LANs, WANs, VPNs – everything.

After working with the product for a few hours though, one has to wonder seriously if anyone outside of the development team ever gave input into its screen design, data-entry fields, message texts or status indicators. Dealing with the GUI was exasperating.

Even setting up a simple replication between two local servers required some unnecessary guesswork. Some examples:

When setting up the parameters for the source and target, there is a grayed-out field that reads “Select Share.” Yet, when you specify a machine that has shares defined, the field remains inaccessible.

Just below it, for both source and target, is a field labeled simply: “password.” Password for what? Help is no help, saying, in essence, “if a password is required, enter it here.” One might guess that it means an Active Directory password but there is no field where a user ID can be entered. Exasperating.

One of the “advanced” features the product claims is “bandwidth throttling.” Yet it, too, is exasperatingly enigmatic. It reads: “Throttle connection at x%.”

Throttle WHICH connection? I know I’m hooked up to a Fast Ethernet network that traverses an asymmetric DSL access across the Internet ending up on a T-1. I also realize that the software can, at best, be aware of the Fast Ethernet.

So, even if I put in 1%, I could theoretically by commanding my source to pour data out at a rate that could consume most of my T-1 if, as I believe, it is a percentage of the local LAN connection that is being requested.

Yet, when I tried to run this “job” using 1%, it just hung. In fact, its state has been “initializing” for two days now. I’d cancel it except I found out that the program doesn’t implement such a command.

Many of these flaws could have been eliminated had ease of use truly been a concern of the vendor. Nothing drives up cost of ownership like a poor user interface.

And all of this leaves me wondering whether the product is a “gem in the rough” or a piece of junk that has such architectural flaws that no user interface upgrade would make the product suitable for an enterprise environment. I shudder to think of having this program be the cornerstone of backup for any truly mission-critical data.

Fortunately for customers, the flaws reveal themselves early into a test drive of the demo version – before any real commitment has to be made.

But time for investigation is a scarce commodity in today’s corporations. Buyers short on time who purchase such products on impulse are in for a rude awakening.

So despite all the praise vendors heap on themselves about ease of use and total cost of ownership, it still remains caveat emptor.