Rent vs. Buy

Saul Hansell's article in the NYT, Why Are iPhone Users Willing to Pay for Content?, tells the obvious part of the story:

Apple has created an environment that makes buying digital goods easy and common. With an infrastructure that supports one-click purchases of songs and videos, it was easy to add applications in the same paradigm. Paying for software, especially games, is not new to Apple customers. So when you see the iPhone manual or the Frommer’s Paris guidebook, it feels natural to click. (And of course, your credit card is already on file with Apple.)

But Lauren Rich Fine's article in PaidContent, Micropayments? Won’t Work, makes the crucial, non-obvious point:

When someone buys a song, that person keeps it forever and plays it many times. News is fleeting, it changes by the second.

I would take it another step. People are clearly willing to put up cash for purchases that seem like one-time investments. Buying a $2.99 iPhone app that I end up using for a couple of days is far more palatable than paying $1/day.

How to Levitate

I was reading Bill Tancer's book Click last night. I generally hate this sort of book. So practical. But Tancer's smart and had access to a ton of data, so whatever.

At one point, he lists the top ten "how to" queries in the US for the four weeks ending 12/21/2007:

1. How to tie a tie
2. How to have sex
3. How to kiss
4. How to lose weight
5. How to write a resume
6. How to levitate
7. How to draw
8. How to get pregnant
9. How to make out
10. How to make a video

He then completely ignores this truly interesting and unexpected insight into the American psyche: that we all wish we could levitate.

I learned to levitate from Henry Sugar, natch.

i2pi

My old friend Josh Reich finally launches out on his own with i2pi, a data consultancy. I've worked with Josh for some five years now and he has the most analytical firepower of anyone I've ever met. Plus he has some really out there geeky hobbies that never cease to amuse me.

If you have data, need data, want to make sense of data or just plain like data, Josh is the person to call.

When I Was

Listened to the Kevin Kelly econtalk podcast the other day. Eh. One thing he said that struck my fancy was from his Edge.org answer to "What Are You Optimistic About?"

Moving back into the past has never been easier. Citizens in developing countries can merely walk back to their villages, where they can live with age-old traditions, and limited choices. If they are eager enough, they can live without modern technology at all. Citizens in the developed world can buy a plane ticket and in less than one day can be settled in a hamlet in Nepal or Mali. If you care to relinquish the options of the present and adopt the limited choices of the past you can live there the rest of your life. Indeed you can choose your time period. If you believe the peak of existence was reached in Neolithic times you can camp out in a clearing in the Amazon; if you suspect the golden age was in the 1890s, you can find a farm among the Amish. We have the incredible opportunity to head into the past, but it is amazing how few people really want to live there. Except for a few rare individuals, no one does.

While only vaguely interesting as an observation, it clicked with my recurring thought that living in the NJ suburbs was like living in the 1950s. This then prompted this list of places I've spent time, and when they are in the 20th century.

1990s	Silicon Valley
1980s	New York
1970s	Los Angeles
1960s	San Francisco
1950s	New Jersey
1940s	Chicago
1930s	Tomorrow
1920s	Miami
1910s	Poughkeepsie
1900s	Boston

I refuse to discuss or defend my choices--especially the scurrilous ones--on the grounds that they are obvious on their face. But I am willing to entertain alternative answers.

Google Shuts dMarc

I never understood Google buying dMarc, or the price they paid. $102 million up front, with an earnout estimated at more than another $1 billion. The company was a sort of automated rep firm for radio, with about 700 stations in their network and selling primarily remnant inventory for low prices.

The idea was solid. There are thousands of radio stations and thousands of advertisers. Creating a clearinghouse would create huge efficiencies, especially if it was an electronic clearinghouse with data that would help advertisers automate their demographic and contextual targeting. There was a chance that dMarc could get some scale and start generating some real revenue. But it was way too soon to tell.

Google's optimism came from their belief that under their guidance dMarc's would grow like AdWords did. But dMarc had a scaling problem Google never had: they employed salespeople to explain their system to advertisers. This was the way it had to be done in the old media: customers were larger organizations used to being catered to, and the sales cycle was long-ish. The advertisers didn't or couldn't be treated like AdWords customers. The 'long tail' and a commissioned sales force are incompatible. Google's optimism turned out to be hubris. Google's attempt to automate the sales effort gutted the revenue and they never recovered.

Google announced they were shutting down radio sales last Thursday.

Why You Won't be Able to Reach Me for the Next 150 Hours

Robin Hanson at Overcoming Bias pointed to an interview he did for EconTalk. I love Hanson's stuff, so I downloaded the podcast and listened to it. It was great. The host kept mentioning other big names he had interviewed, so I went back and looked at the archives. It's a freakin gold mine, even the non-economists:

Eric Raymond on Hacking, Open Source, and the Cathedral and the Bazaar
Shirky on Coase, Collaboration and Here Comes Everybody
Varian on Technology
Chris Anderson on Free
Ayres on Super Crunchers and the Power of Data

There must be like 150 hours of incredibly interesting looking stuff there.

Is Advertising Good or Bad?

[Edit: If you're looking for a less academic and more impassioned view on advertising, check out my more recent post, Advertising, the Fallacy of Perfectibility, and the Best Minds of My Generation.]

Josh's curmudgeonly comment on yesterday's post reminded me of my grandfather saying that advertising is just "a way to convince people to buy things they don't need." This theory of the purpose of advertising is called by economists the persuasive view.

The persuasive view is that advertising changes your preferences in ways not grounded in the actual good being sold. This makes demand more inelastic to price, resulting in higher prices. And since the product differentiation created by advertising is often spurious, you buy one thing when something else--or nothing at all--might have been better. The persuasive view holds that advertising is bad. This is probably the oldest theory of advertising and the most widely held among both economists and non-economists. (I like Kaldor on this view, but Marshall, Chamberlin, Pigou all weighed in.)

The Chicago School, not surprisingly, disagrees. Their view is that advertising is informative: "search costs"--the cost of finding the product that fits your needs--lead to market inefficiencies. If you don't know about better products you can't buy them, so advertisers are doing you a service by letting you know about their products. Advertising is the most efficient way to minimize search costs. Advertising is good. (Stigler is the economist I associate most closely with the informative view.)

Another idea is that advertising is complementary to the product being advertised: Apple has created an image that makes it worthwhile to be seen in the Williamsburg coffee house with a MacBook, not a Dell. This is a component of most luxury goods marketing. Advertising can be either good or bad in the complementary view: you get more from the product but it also costs more, and the net goodness or badness depends on the size of these two effects.

I'm also going to break out another view, because it's not clear to me that it should be considered either part of the informative view or the persuasive view. It's long been believed by economists that companies use advertising as a way of signalling quality: any company spending a ton of money on a Superbowl ad can't be selling schlock (this is akin to the wood paneling in a lawyer's office, or the Persian rug in a banker's office.) If the signalling is of an infomative nature, then it is good. If it is of a persuasive nature, then it is bad. Brand ads are typically persuasive, but could be viewed as informative in their signalling. Kids toy ads, on the other hand, pretend to be informative but are actually using false signalling to be persuasive (false because the quality of kids' toys tends to be inversely proportional to the amount of advertising.)

While the economists can argue over whose view is right, it's pretty clear that ads fall into each of these categories--and often several categories--depending on the product, the audience, the advertiser, the media, etc. Treating marketing as a single thing leads to flawed reasoning.

For instance, when people say that behavioral targeting is "good" because it puts more relevant ads in front of you, they ignore that relevant ads only make your life better when they are informative; when they are persuasive they make your life worse. BT amplifies the societal impact that an ad already has, so trying to (or needing to) justify it on the basis of consumer benefit is meaningless (and a way of escaping the real question: privacy concerns; when in doubt, change the subject.)

If advertising is differentiated in how it affects our well-being, then the question of whether it's good or bad changes. I believe that the best way to build a business is to somehow make your customers' lives better, so the question I'd love to see more startups answering is: how do we increase the good/informative and decrease the bad/persuasive nature of advertising?

You Don't Arb Your Friends

I've always believed that you can't pay people money to watch ads. I also believe you can't pay them to pitch products to their friends, either directly, by endorsing things on social networks or by creating UGC tied to a product-pitch. (I mean, obviously, you can pay them to do these things, but it's not a good business model.)

This belief is probably the most recurring source of disagreement I have with other people in the online marketing business. I used to argue my side by making vague references to the "Attention Economy", but that was just hand-waving. Then I read Looking Good by Doing Good in the Economist. It references Image Motivation and Monetary Incentives in Behaving Prosocially, by Ariely, Bracha and Meier.

The brief:

This paper experimentally examines image motivation--the desire to be liked and well-regarded by others--as a driver in prosocial behavior (doing good), and asks whether extrinsic monetary incentives (doing well) have a detrimental effect on prosocial behavior due to crowding out of image motivation.

Or, as Richard Titmuss noted in 1970: paying people to donate blood reduces the number of donors.

I think this explanation can be made broader: people try to maximize utility, not money. Making money can have a high disutility in some circumstances. This creates asymmetries: someone may work for half an hour to make $5 and then spend that $5 on a magazine, but they won't work for half an hour to get the magazine. That this seems illogical is the reason I've had to make this point so many times, but I've always thought it was pretty self-evidently true. Now I've got science (well, economics anyway) on my side.

The really interesting point, though, is that it is actually cash that creates the disutility. If you want to pay your users, you can use anything except cash.

Self Limbing

Every couple of months I hear a startup pitch how they are going to put the ad agencies out of business. My usual response is along the lines of "have fun with that." The agency business is not especially big or profitable, as businesses go. By definition, the business is a small fraction the size of the media business. And the only way to grow an agency is to hire more people, people that you aren't afraid to have talk to your clients. This is hard.

On the other hand, agencies are annoying. They punch way above their weight. Some junior account exec can make or break your company with the money the budget their client has classed as "Other." Plus, the whole ad agency business is just a bunch of people who sit around all day either schmoozing or being "creative." How inefficient. Clearly a target for a disruptive technology, no?

The merits of this argument aside (I'll address that some other time), the agencies seem to love to assist the process of making themselves irrelevant. I thought of this when I read Josh Kopelman's amusing post on The Giving Tree. I was always in the camp that the loss of the tree was tragic and shortsighted.

A couple of weeks ago, a consortium of various media and technology companies, led by the advertising agency Publicis announced "The Pool", to create online video ad standards. Ad standards allow systems to be created that streamline ads, monitor them, test them, optimize them, target them, etc. These systems promise to create the advertising wonderland of frictionless one-to-one marketing. The friction being, of course, the marketing people.

Most online ad units are standardized. Banner ads were the first to be standardized, and their growth in volume took off immediately afterwards. But the standardization that made banner ads ubiquitous doomed the format to eventual failure: there was very little room for agencies to come up with creative ways to get people's attention. Banner ads are not only not clicked on by users, eye-tracking suggests that they are not even seen. Now marketers can only break through a user's built-in banner-ad shield by putting something insanely relevant in front of them and hoping they look at it by accident.

I understand why Microsoft and Yahoo! want The Pool. The technologization of the ad business requires standards. But Publicis is shooting themselves in the foot. In the end, human creativity is required for display ads to work. Giving in to the engineers' desire to standardize, systematize and streamline means that the ads eventually become ineffective. Ad agencies need to explain to their clients that it is the unusual that gets attention. This is what clients pay them for: the 'unique' in unique selling proposition, the 'big' in big idea.

The media companies may resist, because without standardization they will do less volume. But if they haven't realized that they can't make up the $0.50 CPMs on volume yet, what will it take?

If the agencies want to end up more than a stump to sit on, they need to stop helping the efficiency experts change marketing to nothing more than analytics.

And then We Approach the Strategic Buyers...

Fortune updates my graph of stock market cap to GDP (they use GNP, but the two are pretty similar.) (Fortune link via Mankiw.)


Warren Buffet said "buy at 75%", about where we are now. What I'd like to know is, why 75%?

Looking at the graph, 75% isn't a great place to invest, historically, except at the start of the late bubble, RIP. The great place to invest looks more like 40%-50%. From a valuation point of view, the "right" number should be easier to find than that of any component asset. The inputs and outputs are much more predictable and stable than in a given firm: revenue (GNP), growth (1.8% per year per person), expenses (salaries, government, etc.) and risk (treasury rates.) The only harder thing to predict is the cost of commodities, although we can use futures prices for this to some extent. Run all this through a macro-DCF.

[The other variable that may move the graph is the amount of the domestic economy reflected in the stock market cap number. Not every company, or even every big company, is public. Much of the agricultural and professional services sectors are private, for instance. I tentatively extend as a hypothesis that the IPO boom of the late 90s is one reason for the run-up in the graph in that period, but I haven't looked at a shred of data to back that up, so don't quote me.]

It would be nice to have this model. It would not only be medium-term predictive, but it would also quantify the degree of uncertainty and/or irrationality in the markets It would also be a nice test of the model itself. Any aspiring academics out there?

Cap 'em All

Geithner announces salary caps on executives at companies receiving government money. This is fair and good and, in fact, not an unusual demand when putting money into a company, especially a troubled company. No investor wants to fund a company just to see the money flow back out to executives. No taxpayer should want that either.

I believe this program also applies to the car companies receiving government money. This also makes sense. Any company that gets taxpayer money should have this cap, as a matter of consistency. I hope that companies like Cargill and Archers Daniel Midlands, who receive massive farm subsidies, will be included next time a farm bill is up for debate. This seems only fair, since their executives are paying themselves with our money.

And it's not just federal subsidies we should look at. Any government subsidy should incur the obligation to not pay more than a living wage to those taking the dole. Consider the state and city subsidies of the new Yankee Stadium: nobody in the Yankee's organization should be taking a salary of more than $500k, including the players: I have heard that some are taking home quite a bit more taxpayer money than this.

Preferred tax treatment is no different than government payments, if you think about it. Both result in more money in the company's pocket and less in the taxpayer's pocket. So, executives of not-for-profits and other entities that don't pay taxes should be subject to the salary cap. Charities like the Metropolitan Museum of Art have paid execs more than $500k. Not to mention universities: my alma mater has quite a few people living large on my tax money.

Then there are government subsidies that are not direct cash payments or cash credits. For instance, the use of the airwaves at below-market costs. For the privilege of relying on these, the executives of Disney and GE need to be subject to the salary caps. And the cell phone companies, of course.

Similarly, government enforced monopolies cause higher payments from consumers to monopolies. Because these are government mandated transfers, they are certainly a form of tax. Companies that rely on government enforced monopoly must be included in the salary cap. This gets Major League Baseball, of course, but they are already in there. The cable TV companies are certainly government protected monopolies, so they should be salary-capped. And then there are the patent and copyright holders. Patents and copyrights are inarguably a form of government protected monopoly. So, the movie and record companies, for instance, need to have caps. And then, by extension, the actors and musicians. And don't forget Oprah, and Steven King. Oh, and the software companies.

This is getting to be a pretty broad net, but it's certainly fair to say that if anyone getting taxpayer funded support of their business is going to have a salary cap, then they all should. And if the companies in question don't like it, then maybe they should stop taking my tax money.