The Best. Really?

I read this from Dan Crow, about how Apple has hit its peak this morning:

Steve was famous for his “reality distortion field”. I saw it up close and personal, and it was amazing. But Steve knew that when he turned on the hype, he needed an outstanding product to back it up. The reason he could seemingly bend reality to his will was that products like the iMac, the iPod, the iPhone and the iPad really were exceptional, breakthrough products. Steve’s showmanship was justified

I followed with a discussion with Simon about this issue. He agrees with that statement whereas I mentioned that all Apple has to be is be outstanding enough, where as other companies won’t do well with just being outstanding enough as they lack Apple’s reality distortion field. He then brings up the fact that ‘outstanding’ means a different thing to different people. Which was what I disagreed on.

Coincidentally I was reading Dustin Curtis’ blog post about seeking the best. In the HN discussions, he too brought up that ‘the best’ means a different thing to different people. Again, on this, I disagree.

Wait, the guy who wrote about value multiplicity in markets disagree that “the best” mean different things to different people? What the hell is going on (hint: kinda linkbait)?

I would like to start by assuming there exists such a state of product where it is universally outstanding. Let’s imagine that there is such a thing a Universally Outstanding UX Pattern (UOP). UOP is competing amongst other UX Patterns in Category X. Because UOP is universally outstanding, it caters to every users’ every UX requirement. Ceteris paribus, the rational user would be using UOP instead of other UX patterns of Category X. The allocation would be something like this: UOP – 100%, Product A – 0%. If no participants in the market can make a move in the short run to change the alloation of goods, we can say then the market for UX patterns in Category X is in the core, and that it is pareto optimal.

We can thus define ‘the best’ or ‘the most outstanding’ as an idea or a product or a pattern (or anything really) in a marketplace of ideas/products/patterns where

  • the allocations are in the core for the short run
  • the allocations are pareto optimal
  • there is a clear monopoly of allocations by one idea/product/pattern.

One such UX pattern is the concept of multi-touch smart phones [1]. In the definition above, do note that we use a clear monopoly as a proxy for quality as we assume consumers to be homo economicus.

Of course this is a theoretical model, akin to spherical cows in vacuum moooving at the speed of light. In real life, things aren’t as rosy. Consumer preference sets and individual utility curves come into play. Perception models of a good too affects the utility curves and preference sets. What you end up with is a bunch of noise that makes the signal difficult to isolate. As such, ‘the best’ would appear to be subjective (i.e. subject to each consumer’s preference sets and utility curves).

Signal vs Noise is also a good way to segue into an example of what I mean by the above statement. Jason Fried wrote on the Signal vs Noise blog, that Siri has great interaction design. While Jason hasn’t labelled it ‘the best’ interaction design of an application he’s seen, asking around my office [2], everyone clearly thought it was pretty great, and close to ‘the best’ possible interaction design.

Do better interaction designs exist? Yes. This reddit thread indicates that many people would prefer a different interaction design.

Why then do people think it’s close to ‘the best’? Incomplete information could be one — for example they might not know that better interaction designs exist. Brand preferences (heuristics) could also inform one’s preference sets and thus one’s utility curve. Having a brand preference for Apple products, for example, will make you consider Apple’s products to be closer to ‘the best’ in terms of quality than it actually is.

Here I’m going to introduce the idea of an idea/product/pattern being perceptually ‘the best’ and being actually ‘the best’. This is the point where I disagree with Simon and Dustin on. I believe that there is such a thing (even if mythical) as ‘the best’, but I also believe that there is a ‘the best’ that is subjective to one’s experiences and exposure. When someone says they look at something for the best qualities, I cringe, because to me, they’re looking at the perceptual ‘best’ version of things.

If you are going to go for the best of something, are you going to go for something you perceive to be the best, or objectively the best? If you claim to go for something objectively the best, do you take into account the media factors, or the incomplete information factors, or other factors that change your perception on things, however subtle? Are you able to discount those factors?

I believe that if you do all the discounting objectively and honestly you will come to a terrible conclusion: everything is terrible and nobody is fixing anything [3]. Then you are faced with a choice of your paths to happiness that Herbert Simon called maximizing vs satisficing. Should you maximize or satisfice? I should probably write a blog post about that one day.

And when it comes to running startups, do you think you can get away with just ‘outstanding enough’? Apple in general has a crazily powerful Reality Distortion Field, which can make things from just ‘good enough’ to ‘best’. Can you pull that off? Should you aim for ‘best’? Or just ‘good enough’ and do your best in creating a reality distortion field? What about being lean?

Good luck with your analysis paralysis.

TL;DR – The point I am trying to raise is ‘the best’ is subjective, but there is an objective core to ‘the best’. If you really want the best, you should stip out the subjectivity, and go to the core. Of course you might find out that everything is terrible.

  1. [1]We shall ignore the differences in implementation for now, and just look at the higher level idea of having a multi-touch surface
  2. [2]Of course, the best place for an informal survey. /sarcasm
  3. [3]If you can get out of analysis paralysis that is

Leave a Reply

Your email address will not be published. Required fields are marked *