Dan Heller's Photography Business Blog Industry analysis from www.danheller.com

The photography world -- the business, the culture, the art, the politics, the technology.

Site Feed

Subscribe to
Posts [Atom]

My Photo
Name:
Location: Santa Cruz, California, United States
My Books on the
Photography Business

Wednesday, May 16, 2007

The Photographer's Dilemma: to cooperate or not?

Let's play a game:

You and a friend have robbed a bank. You almost get off scott-free, except for an old lady who "thinks" she saw you, but isn't quite sure. The police round you two up, but because they don't really have enough evidence to convict, you and your friend agree that the best thing to do is not confess anything. Thing is, the police know this too. So, they place you into separate cells, and start playing mind games. The district attorney enters your cell and gives you this to think about:

"I'm giving each of you the same option. If one of you testifies for the prosecution, and the other remains silent, the betrayer goes free, and the other gets the full 10 years. If you both stay silent, we can't convict you on the bigger sentence, but we can get you on lesser charges, giving you six months in jail. If you both betray each other, then you both serve two years."

Each of you has two options: stay quiet, resulting in a lighter sentence for each of you. Or, defect from the pact. The table below illustrates your options:

Prisoner B Stays Silent Prisoner B Betrays
Prisoner A Stays Silent Each serves six months Prisoner A serves ten years
Prisoner B goes free
Prisoner A Betrays Prisoner A goes free
Prisoner B serves ten years
Each serves two years


Because you're in separate cells, you have no idea what your friend is going to do. But here's where your mind plays logic games, and human psychology kicks in. If you knew your friend would stay silent, your best move is to betray, because you then walk free instead of receiving the minor sentence. If you knew your friend would betray you, again your best move is still to betray him, as you receive a lesser sentence than by remaining silent. Of course, your friend will think the same thing and therefore also betray you. Yet by both betraying the other, you both get a worse sentence than if you both stayed silent. So rational, self-interest plays into each of your decisions, making you both worse off than if you'd "cooperated" on the agreement by staying silent.

This is called, "The Prisoner's Dilemma" game, and is the primary basis for many psychological observations of human behavior. And, as such, it is also one fundamental cornerstones of economic theory. The Prisoner's Dilemma (a.k.a. "PD") has been tested in many different contexts, conditions, and demographics since it was originally conceived in 1950 by Merrill Flood and Melvin Dresher working at the RAND corporation. (This is also the basis from which John Forbes Nash derived his famed "Nash Equilibrium", which earned him a Nobel Prize for Economics.) These principles serve as the foundation for most businesses, governments, social policies, and many other aspects of everydy life because they reflect innate human decision-making. So long as there is no interfering body, policing of behaviors, enforcement of compliance, or any other factor to deter this natural human instinct, it is "natural" to serve one's own interests, and also to protect against harm from others by "betraying" them first. In game theory, the Prisoner's Dilemma is called a "zero-sum game" because one wins if another player loses.

If this is the case, why doesn't society collapse? Because we have an intricate and complex system of mechanisms to keep things in balance. We have laws, police to enforce them, social stigmas, and religious dogma, just to name a few. There's also the fact that society isn't necessarily a zero-sum game: people don't kill each other unless there's some gain in it. In fact, there's a great deal of self-protection in maintaining order and keeping peace. Yet, as we all know, there are conditions where people choose to break out of that.

What does this have to do with the photography business? Because, unlike almost every other business, photographers tend to be entirely independent in their business objectives. Accordingly, the way they think and behave is predictable insofar as they will do whatever they can to minimize their own loses, and they also know they can optimize their own returns by betraying their "accomplices," even if it would result in a better payoff for everyone if they didn't. Unless there is a plausible, accepted, and recognized enforcement body over independent photographers--which there isn't--they will, by definition of human behavior, rationalize according to the fundamentals described by the PD game.

This behavior has not only been established in small groups of people, but it has been shown that the greater the number of players, the more assuredly that "defecting" will be a given. It is the First Commandment of economic theory. In any business model, whether it's a sole photographer, or a larger agency selling the work of thousands, this is this fundamental principle that instantly discounts all arguments made by photography interest groups that "sticking together" works. In fact, those who "cooperate" end up serving the ten-year sentences, while the "defectors" are the ones who benefit most. This is so fundamentally basic and obvious, children and lower primates have also shown this obvious and intuitive rationale. (That photo interest groups and educational systems continue to advocate this approach does more harm to those in the photo industry than anything else.)

The implications for photographers have far-reaching effects that go way beyond what most people think. I alluded to this indirectly in my article on keywording, where I stated that the way microstock agencies are set up, they are creating precisely the very conditions of the Prisoner's Dilemma: those who keyword in "cooperation" with how the system is supposed to work will be less likely to have their images found than those who "defect" from the rules and pollute their images' keyword lists with swaths of synonyms, or unrelated (but highly sought after search terms) like "sexy." It is every photographer's objective to be "found," and so the chances that everyone will cooperate with proper keywording techniques is too low to risk the penalties of compliance. So, photographers have incentives to "defect" from the rules.

What's interesting about the state of the photo industry today is that the Prisoner's Dilemma game is going to be put to the test once again with the emergence of user-specified pricing within the microstock arena. We are already seeing this take place in a few companies, such as Fotolia and Lucky Oliver. Traditionally, microstocks have had stable pricing structures of around $1/image (with higher rates for larger sizes). Some sites are beginning to experiment with user-specified pricing--Lucky Oliver and Fotolia let user's extend their prices from $1 up to $100.

While this will certainly present an interesting application of the PD game, there is an even more forceful player coming into the market, if not by sheer size and image base, but because they are are migrating from being a social-networking photo-sharing site, to be being a stock licensing site. Recently, smugmug has alerted their contributor-base that they are within "weeks" of turning on a new feature that will allow licensing of photos. While smugmug has been solely a photo-sharing site for years--and a profitable one at that--it was only in the last year that sufficient push from their membership convinced them to allow them to sell/license their photos. (They had always been able to sell their photos as prints to anyone that wants to buy them.)

In smugmug's new system, photographers will designate those galleries whose images are available for licensing. While the details are still unknown, one aspect of their pricing scheme is that photographers can set their own prices for their works entirely independently of what anyone else does. And, unlike microstock sites that sell everything at a default value, with some collars set on what the user can do, smugmug will have absolutely no bounds whatsoever.

At least, that's the plan for now. The question is, how long will this last before the PD phenomenon kicks in, and pricing is reduced to bare minimums once again, forcing the company to reign in their policies and impose some sort of structure. To illustrate how and why they may have to do this, understand that each player (photographer) has no idea what's really going on in the aggregate sales. That is, the "unknown" element here isn't what another photographer sets his prices at, it's whether he's actually making any sales at that price. It'll be very a difficult psychological challenge for a large set of photographers to price their images at $100 or more, when they see even a small set of photographers with comparatively similar images price theirs at $1, or even less. (Say, $.99 :-)

But, wait. There's more! Not only is smugmug allowing users to set their own prices, but they are joining several other microstock houses in the idea of letting users judge/rate each others' images as well. The theory being that "crowdsourcing" works, and that by cooperation, the aggregate sum of all the ratings will be accurate enough that such ratings can be an integral part of search results. That is, if a user searched for "corporate weenie on cellphone", then of all the vast number of images supplied by hundreds of thousands of users that have such images, the "most popular" ones would rise to the top.

Sure enough, this has been a successful feature of Flickr, who has employed much the same model, which they call "interestingness." Similarly, smugmug and photo.net have seen similar models succeed quite well. There's also amazon.com, where people review books, give them "star" ratings. Users can also build their "collections" of favorite movies, books and music that they share with others. On ebay, users can rate other sellers.

Why doesn't the PD game rules apply here? In short, because there's no incentive to. This is not a zero-sum game, so the psychological factors simply don't kick in. On amazon, the reviewers are not the same as the sellers; one doesn't "win" because someone else loses. Same with Flickr, photo.net or ebay. This is a classic example of how crowdsourcing works really well. (Even though an author can pump up his own book using a fake account, it still doesn't do his book much better by trashing someone else's book, unless they are one of an extremely rare category. And the amazon editors look for this.)

Once you introduce money into the picture, however, the whole system changes. Each of these existing systems would completely transform overnight. And that's why smugmug is more interesting than everyone else who always has been a microstock site. Smugmug hasn't been, but they suddenly will be, making it an interesting large-scale experiment of the PD phenomenon. I'll be looking intently at how it evolves, but I'm not holding out any hopes for optimizing economic returns.

In summary, human nature being what it is, the First Commandment of the laws of economics is, "avoid creating conditions where people's behaviors are rationalized based on the Prisoner's Dilemma." In other words, never set up a business model that depends on the cooperation of those who are inherently competing with one another. It will always fail, unless there is some way to police the activity in some manner. No matter what feature, pricing scheme, or anything else, do not depend on the cooperation of the membership.

So, where does that leave the microstock business model? How do they effectively price their images in light of industry-wide competition? And how can they introduce the benefits of an editorial staff without relying on the user community who have built-in incentives to make the system fail? And how can they build a rating system of their "popular" images if they can't trust the user community to be entirely honest (let alone consistent) about their votes?

To that, I'll have to ask you to tune in next time, as that will be address in my next article.