I have recently acquired a new gadget, an apparently simple device for the conversion of breast milk into kinetic energy, sound and poo. What has become apparent over the past three weeks is how poorly designed this device is from a usability perspective.
Commonly cited requirements for a thing to be considered usable often include that it is simple, learnable, memorable, self-explanatory (intuitive), makes use of existing conventions, not prone to error, satisfying to use, and even that it is beautiful. So where does it all go wrong for my new milk-processing machine?
Well, for starters, it isn't simple. At first glance the controls seem perfectly straightforward: there's a hole for milk to go in, and one for waste to come out. But on closer inspection there are a number of holes, all but two of which (mounted either side of the main milk-entry globe) seem capable of delivering waste in various states of decomposition. Not only that, the milk-entry hole is also the main exit point for the release of sound energy: this seems to be an unnecessary and confusing dual purpose. A small amount of sound energy is also emitted from the main waste exit point, though thankfully less frequently.
As for self-explanatory, I have searched all over the device and there are NO instructions whatsoever. If not for the profusion of existing users and community documentation, discovering all of the features of this product would be a near-impossibility. Very disappointing. Moreover I have noticed that frequently the responses of the device are inconsistent in reference to the user input. Ingress of milk should be followed by a recharging period, and yet this is frequently NOT the case.
This leads me to my main gripe - the appalling error handling. Now I understand that a high-end gadget such as this will require a certain amount of experience to operate to its full potential, but there is simply no excuse for the existence of only a single warning signal. Whether the problem is waste related, a request for input of milk, or something more serious, the warning siren is the same. This seems to needlessly lay the onus of responsibility onto the end-user where the fault-finding process is concerned.
In terms of learnability and memorability, it is true that the fundamental operating procedures are not complex, and even taking the aforementioned error-handling into account, finding the correct procedural response to the notification coming from the device is usually a simple process of elimination. However, the key point of failure here is the fact that this device seems intended - in fact required - to be operated on a less than optimal amount of sleep. This deprivation is taken to such extremes at times that even these simple procedures can seem tiresome and complex - truly a flawed design.
Satisfaction in the operation of the device is currently inconsistent, and rather dependent on how it responds to input from the user. This is something that can be frustratingly inconsistent. Anecdotal evidence suggests that satisfaction levels increase significantly over time.
Overall one is forced to conclude that this device is not merely poorly designed, but seemingly not designed at all. It has certainly not undergone effective usability testing (the shared milk-input and siren hole would never have made it through the first round of testing), or even the most basic forms of heuristic development. I truly cannot imagine what the manufacturers were thinking of, and I suspect that they weren't thinking at all.
All this said, this is still a hugely desirable device, and one that has already won itself many plaudits for the simple fact of being extremely aesthetically pleasing. It is rare that a contraption of such clearly flawed interaction design can still win over users, but in this case it has. I feel confident in saying that the beauty inherent in this creation, together with the promise of increasing user satisfaction, outweighs all of the above flaws.
I just wish it came with an instruction manual and some sensible warning lights.