Friday, November 30, 2007

What is Privacy?

My grandmother used to tell me never to do anything that I wouldn't want to see on the front page of the newspaper. Maybe she lived up to this ethical standard, but she'd be the only one. Substitute "internet" for "newspaper" and here we are.

In today's NY Times, there's an article about Facebook backing off its too aggressive advertising policy. No surprise.

The article had an interesting comment from an exec of a top interactive agency:

Isn’t this community getting a little hypocritical?... Now, all of a sudden, they don’t want to share something?
Well just because they want to share something doesn't mean they want to share it with you. I often have to remind my five year old of that.

More generally, though, the issue isn't revealing personal facts--people do that all the time in every human venue--the issue is having control over the personal facts you reveal.

There is a certain hypocrisy at work, but it's not a 50-million-strong mass hypocrisy. It's the hypocrisy of the media companies that back off personal targeting when it is exposed to their users but keep doing it when it's "behind the scenes, where consumers do not notice it."

Thursday, November 29, 2007

Into the Open, Into the Closed

When Fred Wilson said "open is the new closed" he was echoing Seth Goldstein's "closed is the new open", even if it sounds like he wasn't. I had a long incoherent post (what else is new?) about opening a few weeks ago. Here's a summary: every opening exposes something else that's closed.

Opening up a platform allows closed apps to better utilize it. The exposure of new closed opportunities creates new companies, creates new markets. This is good. It commoditizes the open layer--lowering prices--while creating the monetary incentive to innovate in the new closed layer.

There are two ways to play the opening: creating the open and riding the open. Seth rides the open by building a company that creates, distributes and monetizes closed widgets. Fred and USV creates the open by funding new, disruptive, businesses that commoditize the previously closed. Bug and Clickable are two good examples of companies that won't earn the rents afforded the closed but might garner smaller margins on much larger volumes.

Monday, November 26, 2007

Matchbox Branding

Kids' toys are an interesting business. The velocity of the viral marketing of kids' chatter is exceeded only by that of currency traders' jokes, it seems. One day every kid in town is playing on Club Penguin, the next they're all on WebKinz.

My kids are pretty impressed with brands in an enforcing-important-social-norms sort of way. Playing with the same toys is a way of building community, I suppose, much like listening to the same sorts of music or watching the same TV shows is for adults.

My son got a Matchbox car on Saturday, a Mini-Cooper. How much does Matchbox pay to license the image of the car? I'm sure Matchbox does pay BMW something. But shouldn't it be the other way around? It's the ultimate in product-placement. Every time I look at one of my son's Matchbox replicas of a late 60s muscle car, I get the sort of product lust I never get from ads, left over from the Matchbox cars from when I was a kid. What better way to create product pressure for high-end cars fifteen years from now than molding the brains of five year old boys?

Wednesday, November 21, 2007


Robin Hanson over at Overcoming Bias, one of my favorite blogs, is contemplating stopping his contributions. He says

I've been wondering for a while if I should be blogging. Blogging is less of a conversation than I'd hoped, even among blog coauthors. It feels great to quickly put an idea "out there" in an accessible form, but I'm not sure such ideas have much chance to be built on by others. And it does take time.
Why does it not seem like a conversation? Maybe because Robin's expecting people to comment more?

Fred Wilson, a while ago, said "A blog without comments is a one way medium. And that's not as good as a conversation." I disagree.

If you have something to say, why not blog it and link? Then people can comment on your input in a way that's trackable but is asynchronous. I'll comment when I already have an opinion. But when I need to think about something, I'm not going to go back and find the post and then comment. What would happen to this post if I put it as a comment to Fred's July post linked-to above? I'd be talking to the air.

Professors do research; they also teach seminars. In their research they publish and cite and other researchers read the articles and then respond and cite. In the seminars the professors talk and the students comment. The latter is a great way for the students and professors to learn about the state of play as it stands. The former, though, is the best known way to advance the state of play.

Robin and Eliezer have changed the way I think. I'm not smart enough to comment in real time, but that makes their blog no less valuable. Please don't stop, Robin.

Tuesday, November 20, 2007

Advertising: More of the Same, More More More

I've been puzzling over where behavioral targeting takes advertising for the past few months (i.e. here.) It crystallized for me this morning when I started reading IBM's "The End of Advertising as we know It". (I only read the first page, though, so this post isn't an endorsement of whatever IBM went on to say.)

All of the targeting technologies: contextual, behavioral, social, etc. say they will rid the advertising world of it's primal problem: the half of advertising spend that's wasted but we don't know which half. We will call this Bernbach's Law, because I feel like it.

People have misread Bernbach's Law to say that the half of advertising that is wasted can be recaptured, thus doubling ROI. While this may be true in certain cases--like when you have a CMO much smarter than all the other CMOs--I don't believe it's true if everyone uses the same targeting.

Advertising is two interlinked markets. The first is the media market. Media companies spend money creating content to draw eyeballs and then sell access to those eyeballs. The media business is highly competitive so we should expect that media company pricing is driven primarily by the cost of producing and distributing content. (Both of these costs have changed with the advent of the internet, but we're going to take that as a given.)

The second market is companies making products or services and trying to find customers by letting people know about them through advertising. The cost of marketing is built into the cost of the product, so advertisers are less sensitive to the absolute ROI of advertising as they are to their ROI compared to their competitors. (If everybody has an equal handicap, what that handicap is barely matters.)

What does this mean for the future of targeting? What happens when everyone uses targeting and targeting is perfect, when Bernbach's law is no longer true?

If advertisers could put their ad only in front of the fifty people who are likely to buy the product rather than the hundred that might or might not, those other fifty ad slots go begging. Half the advertising inventory will not be bought. This 50% drop in demand should drastically cut prices.

On the other hand, if an advertiser can target better, she should be willing to spend more to get in front of that targeted audience. But how much more? Twice as much? Does the advertiser end up paying less, resulting in lower overall media revenues? Or does media benefit by being able to charge premium prices for all of their inventory?

Here's my prediction: media revenues will rise overall. Media companies that don't participate in the targeting arms race will lose, of course, but otherwise costs of creating and distributing content will remain the same except that the cost of the targeting technology will have to be added in. CPMs will climb as much as is needed to make up for the fewer ad slots bought.

Advertisers will have the same ROIs that they always have had and the cost of the targeting technology will get passed along to consumer prices.

Consumers have always paid for the media they consume by paying more for products that advertise. Now they will pay more to see fewer (albeit more intrusive) ads.

Tuesday, November 6, 2007

I Don't Mind Being Friends with Brands, I Just Don't Want to Have to be Friendly to My Friends Brand-Friends

Reading the press and opinion on Facebook's new targeted advertising system. I think Facebook is overreaching, they're asking too much. They've given us a container for widgets and a few of their own substandard built-ins (email, pictures, feed.) While free they accumulated the real value: a ton of users' social graphs. Now that they have them, they want to charge (through untrammeled advertising) a huge amount for users to access their own data. The best analogy is probably the old CDDB bait-and-switch.

Seth talks about distinguishing open from closed. I'm not sure why he's confused, we used to talk about this all the time, and he was open's most eloquent dead-philosopher-namedropping proponent. I direct him to go back and read AttentionTrust's principles. Open means the platform gives up trying to control how the user uses it, giving up trying to own the data about the user. Open is the opposite of the walled garden. A platform's implementation can be proprietary, but that doesn't mean it's closed. Open is better for the user, it means giving the user control over what they do and can do. Open is better for the platform as well, it means allowing the user (and others) to innovate and then being able to spread that innovation to other users.

We've been given small taste of openness. Now I can't do without it. So I want to know: where's the OpenSocial portable social graph widget? Can we move that to the top of the priority list?

Here's what I want: a PSG widget that I can import into my NetVibes page that will let other authorized widgets see my social graph (or a specific subset of it) and can communicate with my friends' PSG widgets (embedded in any OpenSocial platform), so they can exchange feed-items. I don't want to be beholden to a single widget container, I don't want anyone to be able to hold my data hostage.

John Anderton, you could use a Guinness right now.

I've got a lot on my mind. You can see it by reading about the weird variety of things I've posted about in the last three months. For the most part it's about the future of marketing. Much of it is, uh, aspirational (read: wishful thinking.)

Greg posts about behavioral targeting: he's tired of it. I think he means he's tired of hearing about how it's the new marketing revolution. Me too. It's not new. Behavioral targeting is what marketing has always been, we're just better at it now.

I don't like behavioral targeting. Not because I don't think it effectively sells product. I don't like it because I think it's destructive of identity. It's interesting that, on the one hand, we have companies like Facebook flourishing because they protect the privacy of the social graph while, at the same time, marketers undermine that privacy to sell things.

If there's one thing the explosion in communication forms (or even the popularity of Twitter) proves, it's that our identities are mappable to our relationships. Relationships require trust, trust requires intimacy, the willingness to be intimate requires the ability to enforce privacy. Without the ability to enforce privacy, we won't form real relationships.

I'm waiting for the first viable social network whose technology does not allow it to violate the expectations of privacy that Facebook built itself on and is now jettisoning. I think it will take the form of individual open-source widgets embedded in a web page that is nothing more than a way to regulate the communications between the widgets. Anyone?

Sunday, November 4, 2007

Keep Your Content Close, but Your Friends Closer

Would you rather spend your time finding the needle in the haystack or the haystack on the needle?

Randall Stross writes an interesting article about OpenSocial in today's New York Times. I'm still muddling my way through where I think the widget ecosystem brings the internet, but Stross says something that contradicts one of my earliest conclusions.

He quotes Joe Kraus from Google saying that the point of OpenSocial is to allow social networks to be mobile: you could embed a widget exposing your network of friends into Craigslist, for example, porting your trusted community into your ecommerce. Interesting.

But backwards. It's far more likely that you will embed a Craigslist widget in your social network page than vice versa, in my opinion. Facebook profiles are home pages, people keep them open in the background while they work so they can keep on top of what their network is doing. It's also far more likely that your friends will see your Craigslist posting if you embed it on your homepage than someone searching Craigslist will run across something posted by a friend.

Okay, its not a dichotomy like that. I know there will be hybrid models. But the fundamental question is: will social features be embedded in content sites or content features in social networks? I'm tempted to say there is a matrix of networking and content features and that there will be opportunities at each matrix juncture, but that's not how it works in real life.

I think social networks become the predominant form of navigation and that pointers to content get distributed through them. This is what is happening--with increasing momentum--now. Why Mr. Kraus and Mr. Stross thinks otherwise is a mystery to me. But I'm certainly open to the possibility that I'm wrong.

Thursday, November 1, 2007

The Sense in Irrationality

There's a more cynical view of why it is relatively unusual for businesspeople to go back on their word than the one I posted about a couple of days ago. I mentioned that the cost of redress when treated unfairly often made it economically irrational to try to enforce a broken contract. I said that more contracts aren't broken because people are naturally trustworthy.

Another reason may be that people are often quite economically irrational.

There's an experiment in behavioural economics known as the Ultimatum Game. This game gives one person a certain amount of money and asks them to split it however they want with a second person. If that person accepts, then both people get the money as allocated by the first person. If the second person refuses, however, neither gets anything. The game is played once, anonymously.

The rational strategy for the second person is to accept any split that offers them anything. The alternative is to get nothing. Thus, the rational strategy for the first person is to offer the minimum amount to the second person.

This is not what happens. Although it varies from culture to culture, any split offering less than 20% to the second player is usually rejected. People are willing to act somewhat irrationally to enforce an idea of justice. Economists, who often base their theories on the idea that each person is an autonomous unit making only decisions that benefit themselves, can not explain this result. But it's easily explained as a group dynamic: if everyone was willing to personally accept some cost to punish the unjust, there would no longer be any injustice. It wouldn't pay.