Monday, July 30, 2007

Ozzie at Microsoft

As Bill Gates prepares to step down from Microsoft, it seems that his role will be taken over by Ray Ozzie and Craig Mundie. I don't know much about Craig Mundie, but I've seen Ray Ozzie's work when I interned at Lotus, and frankly it was not good. I was also deeply unimpressed by Groove (as was Joel Spolsky).

Friday, July 27, 2007


I'm not sure if you're reading Marty Cagan's excellent Silicon Valley Product Group newsletter, but if you're not, you should.

This great article on the iPhone is one example of why. I agree with everything Marty says -- particularly his point about focus groups being useful -- but I'm not sure I agree with his final comment:
I’m not sure about the positioning of this device primarily as a phone. The device is much more than a phone, in fact, I’d rate the fact that it’s a phone (and I consider it the very first truly good phone interface out there) as secondary. The device is much more of a complement to the personal computer. It’s a bit like bringing my Mac with me everywhere I go. Perhaps more importantly, it is also the first mobile device that I can imagine being the only computing device many people have.
It is true that the iPhone is more than a phone, but it needs to be a great phone first and foremost before it's going to win that valuable, and limited, pocket realestate. Personally I cannot imagine having *only* an iPhone -- as good as the keyboard and screen is, I have no doubt I can type faster on my laptop. The laptop is a foot scale device, suitable for foot scale tasks. The iPhone is an inch scale device, suitable for inch scale tasks. Both are important and useful, but both are also different.

Cellphone bills

I liked this funny post by David Pogue on his first iPhone cell phone bill.
For starters, although I signed up for what iTunes told me was a $60 plan (450 minutes, unlimited Internet), the bill says I have a $40 voice plan and a $20 Internet plan, and lists them on separate pages.

The first bill, believe it or not, comes to $150. It’s filled with unexplained services and features that were never mentioned during the signup process, like MEDIA MAX, EXPD M2M, VOICE PRIVACY, and AT&T DIRECT BILL.

After studying this thing for 20 minutes, I think I’ve got it figured out: activation charge ($36), prorated monthly fee for June ($26 for voice, $13 for Internet use), taxes and fees ($15), plus July billed in advance ($40 voice, $20 data).

All of that fits on three dense pages. But then–get this–I get SIX PAGES of listings of data tidbits that the iPhone has downloaded in the form of email and Web pages–KILOBYTE BY KILOBYTE! Every graphic on every Web page, every message sent or received–it’s all carefully listed by date and time.
Heh heh heh. Cell phone companies should just wrap their product in a big sticker that says "your first bill will be twice as large as you think". Billing one month in advance totally confuses customers, and phone companies are bad (generally) about explaining it. That means a $60 plan will cost you $120 the first month. David got off lucky with just $15 in taxes -- if it was a landline it would cost $60++. The listing of data consumption is a simple holdover that no one thought about (or had the time to address) -- it's this kind of detail that seperates great customer experiences from regular interactions.

As landline usage shrinks and cell phone usage grows expect to see all kinds of taxes added to cellular plans. Governments want that money.

Thursday, July 26, 2007

Used cars?

I like the used car market, it brings up all sorts of interesting economic and behavioral phenomenon. HotSwap lets used car sellers advertise their cars online by posting videos. On that web2.0 theme, it also features guided nav with color swatches, sliders to refine searches, and lots of light grey text on a white background. I'm sure tags are coming soon.

I like the video as a better way of showing the car, but I don't think it really gets to the information asymmetry issues that are meant to haunt the used car market. I still think that CarMax is the best way to buy a used car if you're short on time and mechanical nous. But HotSwap is a step up from static used car listings from individual sellers.

Wednesday, July 25, 2007

Complements or Substitutes

It's not always obvious whether two goods are complements or substitutes. For example, is a radio play a complement or a substitute for a CD?

An economics professor at UT Dallas has tried to answer this question and reports that they are substitutes, ie. radio listeners seem to buy less music(NYTimes)
It found that, very roughly, an hour’s worth of radio listening per person per day, over the course of a year, corresponded with a 0.75 drop in the number of albums purchased per capita in a given city.
Clearly the cross-product elasticities are more complicated. Radio play undoubtedly helps create hit songs (positive cross elasticity: complements) but I can also believe that having really really good radio may make you buy fewer albums overall (negative cross elasticity: substitutes).

From this, it seems that it's efficient for music owners to pay radio stations to play songs that are likely to become hits (sadly outlawed under "payola" rules) while radio stations should play music owners when playing more obscure music that will likely never become a hit. Maybe it all works out in the wash.

(Interesting related article on new pricing scheme that is likely to shut down web radio [NYTimes]. Bummer -- I like Pandora and

Regulatory Inefficiency Theorem

This funny post from Long or Short Capital highlights the absurd path that telco regulation has taken in the US.
The Regulatory Inefficiency Theorem states that any move by a regulator will thereafter be reversed by the market. The time to reversal will vary inversely with the amount of regulatory oversight implemented whereas the lobbying dollars spent will vary directly.
The article ends by noting that the telecom industry spends more on lobbying “than the tobacco, aerospace and gambling lobbies combined.” If you want to know why it's sensible for them to do this, check out the taxes on your local and long distance landline bill.

Tuesday, July 24, 2007

iPhone dud?

This piece states that fewer iPhones were sold than forecast, which sent Apple and AT&T stocks lower.

It may be still to early to say, but perhaps, for all its wizardry, the iPhone is still not good enough as a phone to convince people to switch. I've always contended that cell phones need to be cell phones first, which means focusing on 1) reception and 2) voice quality. I did not see much in the iPhone's marketing material about 2), and AT&T is not known for 1).

If the iPhone does dissappoint, I'm sure that tech pundits will say it's because it's a closed platform, without any third party applications. I think the real reason will be that it's missing some core phone basics, and because it's too expensive for something as fragile and losable as a cell phone.

Friday, July 20, 2007

Hal Varian at Google

Berkeley economist Hal Varian is now at Google. Varian's "Information Rules" book was very good -- and fairly evil.

Wednesday, July 18, 2007

Treating the sick

This Slate article does a good job of drawing attention to the plight of the very sick in poor countries, but a terrible job of thinking through the issues. The standard way to pose the problem is:
If you have $10,000 to spend on healthcare in a poor country, is it better to
a) operate on one person with a bad heart, giving him an extra 6 years of life
b) invest in childhood preventative care for 100 children, giving each one an extra 5 years of life

The mathematics is pretty compelling: you get 6 years for your $10,000 in one case, and 500 years for that same $10,000 in the other. That is if you believe in this sort of medical mathematics. The Slate article does not (a position I have some sympathy with) and rejects the whole notion that you can value human life at all.

The rest of the article has numerous examples where very expensive treatments have been given to people in very poor countries. This has been done either through charity (doctors donating time, pharma companies donating drugs etc.) or through breaking intellectual property patents (which I also have some sympathy for).

Neither of these approaches though do anything to confound the mathematics that cheap prevention across hundreds of people is more cost effective than expensive interventions in just a few cases

Wednesday, July 11, 2007


Just heard a news story on NPR about the hyperinflation and economic meltdown in Zimbabwe. "Robert Mugabe" was not mentioned once, although transfers of money from foriegn taxpayers was stressed many times. Sheesh.


This poor piece on Slate applauds Sen. Herb Kohl, D-Wis bill which requires the Justice department to take on OPEC becauseOPEC is a cartel and cartels are illegal.
Kohl drafted a bill, dubbed "NOPEC," that said OPEC could no longer protect itself from antitrust prosecution by citing "sovereign immunity" and explicitly granted the Justice Department jurisdiction. The bill went nowhere back in 2000. But this past spring, Kohl dusted it off, and John Fialka reports in the July 6 Wall Street Journal that NOPEC has won the support of veto-proof majorities in the House and Senate. The appeal of NOPEC extends from left to right; House Speaker Nancy Pelosi, D-Calif., is pushing it, and so is the Heritage Foundation. The Bush administration, however, can't stand the idea.

The White House Office of Management and Budget says it opposes the NOPEC bill "adamantly." Perhaps this is because, as I've noted before, OPEC is just about the only international organization that President Bush has any regard for. (Shortly after taking office, Bush lectured a reporter about the importance of keeping oil prices "stable and predictable," even if that meant preventing a price drop.) Conceivably Bush is worried that busting OPEC might give Russia too free a hand in setting oil prices. (Russia is not an OPEC member, and its oil production now rivals, and may actually exceed, that of Saudi Arabia.) But busting OPEC would weaken two of Bush's least-favorite regimes, Iran and Venezuela. The Saudis wouldn't be happy, but neither would they be terribly impoverished, given the significant power they'd retain (as guardians of one-quarter of the world's proven oil reserves—about four times those of Russia) to affect prices. Diplomatically, busting OPEC strikes me as a wash at worst.
The thinking here is so wrong, it's hard to know where to start.

Firstly, OPEC has been a failure at setting prices, at least for the last 30 years. Like in any cartel, there is an incentive for each member to cheat, and no effective way to punish cheaters. And as predictably as the rising and setting of the sun, this has led (and will continue to lead) to OPEC members cheating on each other, producing more when prices are high, and thus bringing prices down. Therefore OPEC has no influence in setting oil prices, certainly as production in non-OPEC countries has risen--trying to "bust" it will have no impact on anything, and wastes taxpayer money. (Actually, this is an improvement from what Congress usually does, since it has zero impact).

Secondly, even if the Justice Department tried to break up OPEC, what would it do? Tell the Saudi government and the Iranian government to stop talking to each other? And if they refused, would an American judge tell them to go to a comfy prison in Brussels? I would actually like this to happen, because it would be funny.

As there is clearly no enforcement mechanism, it's reasonable to conclude that Sen. Herb Kohl, D-Wis is grandstanding because it will help him politically. So instead of calling a spade a spade, why does Slate's Timothy Noah applaud the effort?

Monday, July 09, 2007

Responsive to shocks

Many moons ago, Steven Den Beste complained that "a-list" bloggers (like were hogging the limelight, and Clay Shirky was talking about how blog readership followed the power-law distribution, with just a few sites accounting for most of the eyeballs.

All of this was to suggest that the Internet, far from being a fluid medium, was ossifying into the old privileged hierarchies we see in the real world.

Then 9/11 happened, USS Clueless was catapulted into the limelight (as Steven knew something about Apache helicopters, and did not). The powerlaw turned out to reflect real (shiftable) interest, and not ossification. An equal opportunity world does not produce equal results.

I get the same feel from Unqualified Reservations that I used to get at USS Clueless before he became big.

Monday, July 02, 2007

Wet pavements cause rain

The NYTimes has a piece drawing a parallel between the UK's unfortunately named "ring of steel" and a walled fortress. The "ring of steel" is a network of barriers, checkpoints, and most importantly, video cameras, that track movement within central London
FOR more than a century now, winged dragons flanking a shield have guarded each entrance to the City of London. In recent decades, this coat of arms has been reinforced with an elaborate anti-terrorism apparatus known as the "ring of steel," consisting of concrete barriers, checkpoints and thousands of video cameras. City planners call the system, set up to defend against bombings by the Irish Republican Army, "fortress urbanism."

Perhaps no city in the Western world is better equipped to deter terrorist bombings. Yet the two waves of attacks this month have demonstrated that in London, "fortress urbanism" is far from impregnable.

Like the simple wooden ladder that was used to circumvent a castle's stone walls and moat, determined terrorists can still find tools to strike almost at will, even if their plans do not always succeed, as apparently happened last week in London.
This seems to be the exact opposite description of reality. As I wrote recently in "moats vs policemen" the only alternative to raising barriers and becoming a fortress is to let everyone in, but then carefully watch them to find wrongdoers. Instead of classifying surveillance cameras as part of a "fortress" strategy, the NYTimes should instead characterize them as what they are -- the only credible alternative to a true "fortress". (A surveillance based approach is also, IMHO, more likely to be successful).