This is the question I started out with. User experience is something I consider every day when trying to get people to do things they have no interest in, like give my company money, but can it go so far as to get a cat to do something he has no interest in – like caring for a houseplant? Fundamentally, he’s still a user, right?
With this assumption, I began the experiment. The key to user experience design is, predictably, learning about your user. A quick look through his few worldly possessions gave me just the clue I needed: a cat fishing rod with a dangly toy mouse on the end.
Now for the engineering
By attaching the cat fishing rod to a tripod, and tying a shoelace between the handle of the rod and the watering can, I created a simple lever system. In theory, when the cat pulls on the mouse, the lever would raise the handle of the watering can, tilting the spout toward the plant. Water should then pour out, quenching the thirsty soil below. In theory.
The importance of usability testing
Before actually putting water in the can, I had to run some tests. Since I’m not my user, running those tests myself gives me an inherent bias. I pulled on the mouse myself and it worked fine – but would it work when my user did it?
I called him over dangled the mouse in front of him. He went for it, just as I’d hoped – but being a cat, he pounced suddenly with all his strength, and the can went flying. If there had been water in it, it would have been a UX fail as well as a PR disaster.
Time for my first round of revisions
With some kitchen supplies and a bit of copper wire, I had finagled a much more stable system. At the request of my spouse, it was even more foolproof than it might’ve needed to be.
Ready for another testing phase. My user was able to easily put the empty watering can through the proper motions. Excited to see this work, I added the water to the watering can and had him go at it again.
Nothing happened. The added weight of the water and the flimsiness of the cat fishing rod were colliding in a perfect storm of unexpected fail. I needed a better lever – and luckily I happened to have a sturdy wood dowel lying around. (Different project. You’ll see it soon.)
This was the last round of revisions. The improved lever made it much easier to lift the watering can by pulling on the mouse. One problem – my user had gotten bored and decided to take a nap. As it turns out, 3pm on a Sunday is a poor media buy for reaching a feline audience.
Advertising goes bad fast
At 10pm he was back in full swing. The mouse, however, was still of little interest to him. This is what happens to any ad that runs too long, pops up too often, or simply doesn’t offer anything interesting, useful, or novel – it becomes a mere annoyance at best, and ignored completely at worst.
I needed a re-skin. What else was my user interested in? Treats, for one. Tying a small fish-shaped treat to the mouse generated some attention, but not enough. Thus is the fate of the gimmick.
What else could I do? What was the one thing my user would go totally nuts over? Ah-ha! Catnip! I rubbed some on the treat-laden mouse, sat back, and recorded. (I brushed off some of the catnip flakes that were stuck to my hand into the plant itself, hence his initial interest in the plant.)
Success! He was a little crazy for an hour or so before the come-down, but my plant was watered and I had learned some important lessons about getting people to do things they don’t give a crap about. Here they are:
1. Always test. And then test some more. And then probably a third time.
2. Study your user. If you can actually search through his stuff, that’s best. But if not ending up with a restraining order is important to the success of your project, try doing some interviews.
3. UX alone isn’t enough. You need engineering to make it function, advertising to catch attention, good timing to catch the user when he’s most receptive, and, if at all possible, drugs.