Thursday, June 09, 2005

Einstein's Refridgerator, French Maid Costumes, and Captain Obvious

Whether you love Paul Graham's essays or you hate them, you have to admit they're thought-provoking. The first part of this post's title comes from an evocative image in one of his recent essays, illustrating the idea that startups can make insanely great products by getting really, really smart people to think about problems that people actually care about solving:
That's the essence of a startup: having brilliant people do work that's beneath them.... Think Einstein designing refrigerators.
Graham argues that one of the hardest things to do in starting up a company is coming up with something that people want. Sadly, it is even harder in some ways than he lets on. The trick is in that word "want". You see, for hackerly types, there's a natural (to us, anyway) assumption that "want" is connected to a rational evaluation of what is "good", or would be good. This then leads to the kind of interesting place where you can make a product that's really good for your customers, but which they don't necessarily want.

The first startup I worked at was one that made this mistake. It wasn't a fatal mistake by any means; our company survived under the same ownership for longer than any other company in our vertical market niche industry; we watched numerous companies come and go that had more money, better marketing, and crappy products. So, the fact that we had friendly, helpful service, and a product that actually benefitted people meant we got word-of-mouth that kept us alive for quite some time. This only reinforced my belief that the way to ultimate success was creating a quality product that helps people do things well.

Too bad it's completely wrong. Yes, quality helps. But the real reason we outlasted our competition (I think now in retrospect) is that any sane competition got the hell out once they realized that management tools aren't what real estate people wanted to buy. Sure, they could be sold the products, easily enough, and if you were willing to spend time educating them about why it was a good idea, they might even use them.

But education is expensive. It takes time. It can't be mass-produced -- well, actually it can. We eventually figured out that we could videotape our education/sales pitch, and were able to automate our sales pipeline to a certain extent eventually, where sales consisted mostly of calling people and hounding them to watch the videotape. But it cost an enormous amount of roadwork and "sales R&D" to reduce it to such an automatable state in the first place -- costs that wouldn't have been there if we'd just sold something that people already wanted in the first place.

And this is where the next part of this post's title comes in. Halloween is the biggest part of the sales year for my wife's lingerie/adult store, and she sells a lot of costumes. Earlier in the year, often even earlier than this, she goes through the costume catalogs to figure out what to buy. Every year, the costume companies release lots of new and innovative costume ideas -- that mostly don't sell. What does sell? French maids, harem girls, policeman/woman outfits, fairytale characters, and recent movie characters.

(By the way, what I mean when I say that something "sells", I mean that it outsells other products N:1, not that other products never sell, or that only obvious or stereotypical producs are ever purchased. To a retailer, when something "sells", it simply means that people purchase it at some steady rate, as opposed to products that somebody will buy, but you're never sure when or how many. Products that "sell" are essential to having a viable business.)

Anyway, the list of costumes that "sell" is actually longer than the one I gave above, but I don't remember all the others just now. I do know that if I could name them, you would all nod your heads and say, "Of course," because they're all "classics" in the sense that there's this cultural stereotype or archetype that the costume reflects. I don't say this in order to try and delve into the psychology of wearing costumes or anything like that, though. My point is much simpler: people buy what they already know they want; they buy what is familiar; they buy what they recognize, even if the only thing they have to recognize it from is a vague idea in their mind.

For example, my wife's store sells a number of "romantic games", including one called "Let's F*ck" - and I didn't add the asterisk, that's just what it says on the box. This game sells quite well, as does one called "Ultimate Sexual Fantasies". There are lots of other games, and if you study their names for a while you can tell why some sell and some don't. The ones that do, have names that are recognizable -- when the customer reads it, it's something they recognize as "what they want,"even if they never actually heard of that game before. Games with names like "Let's F*ck", then, work surprisingly well.

In marketing this concept is sometimes called "Overt Benefit". It basically means that you tell the customer what it is that they're going to get, and you do it so overtly that to anybody else in your industry it sounds like you're bloody belaboring the obvious. The more insanely obvious you can be, the better off you are. If your game has a less overt title like "Postman's Promise", then you're going to have to spend more effort communicating your sales message, unless that message is at least as overt as "Let's F*ck".

To further illustrate this principle, let's look at Gilligan's Island. More specifically, the new season of TBS' reality show based on Gilligan's Island. They held a contest to decide which of the two "Skipper" candidates would get to stay on the island. In this competition, two teams of "castaways" sat in boats, that then had holes opened in the bottom to create leaks. They would then have to try to plug the leaks and bail the water out. The Skipper whose boat remained above water longest, would be the winner. (In order to prevent a stalemate, the team members were called on one by one to leave the boats, so that fewer people were left to cope with the ever increasing number of leaks.)

The two Skippers handled the situation very differently from each other. One simply sat in the boat with his teammates until the contest began. The other spoke softly -- but urgently -- to all his team members. He told them he wanted them to plug as many holes as they could with their body parts, and to make sure that the coconuts were full of water when they threw the water overboard.

I imagine you can guess which team won, even though it makes no rational sense. Obviously, if the goal is to keep the boat afloat, you should plug as many holes as you can, and bail a lot of water. In fact, it's hard to imagine how he could possibly be any more more obvious! (As opposed to simply being more detailed.)

But the mood of the two boats was completely different. The team belonging to "Captain Obvious" was calm, cool, collected. They co-ordinated with each other and kept their boat surprisingly free of water. The other group flailed around ineffectually, almost randomly. They were bailing harder and faster than the "obvious" group, yet their boat continued to sink. Why? Because they weren't blocking as many holes -- perhaps because they had not been told to plug as many as they could!

Now, it's just an anecdote, but it illustrates a point about both leadership and marketing. In both fields, there is no such thing as what people "should know". If you didn't tell them, it doesn't count. If your product isn't as clear in its benefit description as "Let's F*ck", or if your instructions to your people about what's important (and what they should do about it) aren't as clear as those of "Captain Obvious", then your boat still might not sink -- but you will be bailing a lot more water than would otherwise be necessary.

It's easy, by the way, to think that this doesn't apply to you, that you have a much more rational approach to purchasing or anything else. This is an illusion, because every person has areas of their life in which they are skilled or possess special knowledge, and in that area, their purchasing or other actions may approach rationality. They then use this example to presume their behavior is rational in areas where they don't actually know much about the subject, or in circumstances where they need to fix a problem quickly, etc. I'm probably great at rationally evaluating a software purchase, but if I need car parts or something like that, the marketer is going to have to assume I've got attention-deficit disorder, because I really am going to be focused on getting my car to work, not on what parts are "better" in some absolute sense.

And so, this kind of state actually constitutes the bulk of people's buying behavior, where they will choose whatever it appears will obviously get the job done. The problem, then, is not so much that people are irrational, as that they are generally far too busy to pay attention to details. If your idea or product requires careful attention to understand its benefits, you will lose out to Captain Obvious.

Similarly, managers often fail to tell their teams what they want with sufficient clarity, because they assume that what's obvious to them is obvious to their team, and that "as professionals", they "should just know". But this isn't the case at all!

A long time ago, I read that researchers have shown that programmers will optimize whatever you tell them (implicitly or explicitly) to optimize. If you emphasize clean code, you'll get clean code. If you emphasize features, you get features. If you emphasize speed, you'll get that. However, if you don't tell them anything, or you give them conflicting directives, God only knows what you'll get, but it won't be pretty. So remember: if you don't sound to yourself like you're belaboring the obvious, you're almost certainly not being obvious enough.

Have I made my point yet, or do I need to belabor it some more? :)