How Much Apache?

Almost with inexorable momentum, the Internet hurls itself into new territory. Some time ago, more than two billion humans had adopted at least one Internet-enabled device in some form, and nobody doubts that another two billion will accrue soon. New webpages increasingly find ways to inform readers, as more information in a variety of formats continues to be layered on the basic system of data internetworking.

That growth has been measured in a variety of dimensions. Today I would like to report on some research to measure one aspect of the Web’s growth, which I did with Frank Nagle, a doctoral student at Harvard Business School. We sought to figure out how much Apache served web surfers in the United States.

That is not a misprint. Apache is the name for the most popular webserver in the world. It is believed to be the second most popular open source project after Linux.

Why do this? Measuring Apache is a key step in understanding the underlying economics. Because it’s free, Apache’s value is easy to mismeasure, and that makes its economics easy to misunderstand.

End the broadband panic meme

 

It happens about every twelve months, maybe with more frequency recently. Another reporter writes about how the US is falling behind international rivals in the supply of broadband. I am growing very tired of this meme, and answering emails from friends wondering if it is so. There are serious issues to debate, but this standard meme takes attention away from them.

 

The latest version of this article came from the New York Times. It had the title “US Struggling to Keep Pace in Broadband Service,” and it brought out the usual concern that all US growth will fall behind if the US does not have the fastest broadband in the world. If you are curious, read this.

 

Why is this tiring? Let me count the ways.

 

First, while it is irritating to have slow service at home, US productivity does not depend much on that. Household broadband is less important for economic growth than the broadband to business. And what really matters for productivity? Speed to business. The number of minutes it takes a household to download Netflix is statistically irrelevant for productivity growth in comparison to the time it takes to download information to conduct business transactions with employees, suppliers, and customers. We get measures of broadband speed to homes because that is what we can easily measure, not because it really matters.

 

Is there any sense that US business Internet is too slow? Well, perhaps the speed of a household’s internet says something about the speed of business Internet, but I doubt it. In all the major cities of the US there is no crisis at all in the provision of broadband.  Broadband speeds in downtown Manhattan are extraordinary, as well as in Wall Street. The Silicon Valley firms who need fast speeds can get them. Same with the firms in Seattle. Hey, the experiments with Google Fiber in Kansas City raise questions about whether entrepreneurship will follow the installation of super high speeds, but that is an open question. It is an interesting question too, but not a crisis.

 

These issues do arise, however, in some small and medium cities in the US, and a few rural areas where there is no broadband. In some places satellite is the best available, or some fixed wireless solutions are available too. These can be ok but not great for many business needs, but it can also limit what a business can do. These issues also have been present for a while, so most of the businesses that really needed the speed simply left the areas where speeds were slow. As a country we just let it happen a many years ago, and, frankly, it will be hard to reverse at this point. (It made me sad at the time; I even spent some time doing research on the topic for a while, though I have stopped in the last few years.) Again, this is an interesting question, but only a crisis in the places where it matters, not a national level.

 

Second, as for household speeds, many people simply don’t want them and do not want to pay for them. There is plenty of evidence that those high speed Korean lines did not get used right away, and lots of fiber goes to waste. Having said that, there are some interesting open questions here as well, namely, what type of speeds are people willing to pay for at their homes? Let’s not get panicked over supply if there is little demand, ok?

The last serious study of the willingness to pay for speed was done at the end of 2009, as part of the national broadband plan. The study was definitive at the time, that only a few households were willing to pay for high speeds. But, of course, that was a while ago. What has changed since then? Well, arguably, demand for data-intensive stuff has risen. That is not coming from the growth in torrent. Recent data are pretty clear about that. It is coming from Netflix, YouTube, and Facebook. Once again, that is a great open question, but panic about speed does nothing to focus on that question. Instead, let’s study demand and whether it goes unsatisfied.

 

Third, if we study demand, can we all acknowledge that demand is very skewed in the US? 10% of the users account for far more than 50% of the data to households, and 20% of the users get most systems to more than 80% of the data use. And it is growing at levels from median to highest part of the skew, so there is good reason to think demand for data is growing for all major users. Will there be capacity to handle those intensive users of data? The answer is unclear.

 

That hints at an open question that is worth debating. Not everyone pays the same price because flat rate pricing has been so common across the US. The top 10% of users pay very low prices per megabit. Even if total expenditure per month for the biggest users is twice as expensive in the US in comparison to other countries, it is still pretty cheap. Just to be clear, I am not saying it is too high or too low, nor am I am not making any comment about whether markets are competitive enough in the US. I am just saying that the international comparisons are flawed for big users in the US.

 

That hints at an even more challenging question. For better or worse, it is these high-intensity users, especially many with young adults or teenagers, who seem to be the early users of new services. So US entrepreneurial edge might actually be coming from the low prices and high speeds our biggest users have enjoyed all these years. Are we in danger of ending that? That is the provocative question to ask, and it is not about the general speed in the country. It is about the highest speeds to select users.

 

Finally, and my last problem with this meme: it’s old and tired and potentially irrelevant. Maybe this concern about wireline is all a tempest in a teapot. Many observers believe wireless is the new frontier for innovative applications. Maybe five years from now everybody will look back on this panic and just shake their heads. How can we have an entire article about broadband speeds to households and not a peep about the experience most people have on a daily level, which is determined by wireless speeds?

 

Just something to think about.

 

Top Dozen Tech Events of 2013

It is time to look back, and give some awards for the best events in information and communications technology. And what a year it was — with Snowden, Healthcare IT, the Twitter IPO, and plenty of other events deserving both recognition and sarcastic observation.

Just like last year, there are four criteria for winning. The winner had to do something in the calendar year. The action had to involve information and communications technology. It had to be notable. That is not asking much, so the final feature is the most important: The award winner has to contain something that deserves a snarky remark or a bit of sarcasm. Like last year, every winner gets a virtual trophy called a “Sally,” affectionately named for Sally Fields. Why her? Because she memorably said, “You like me, you really like me.” That label is meant to convey a simple message: none of this should be taken too seriously.

Here are a dozen. If you disagree with my choices for awards, feel free to suggest your own in the comments section. Let’s get to it.

William C. Lowe

All students of the computer industry have heard of Bill Lowe, the leader of an IBM Boca Raton facility that launched the IBM PC. That launch was a signal event in computing. It catalyzed growth in the small systems market.

I had the great pleasure to interview Bill a few years ago for a research project comparing the response of large firms to external events (Tim Bresnahan, Shane Greenstein, and Rebecca Henderson, 2012. “Schumpeterian Economies and Diseconomies of Scope: Illustrations from the Histories of IBM and Microsoft,” The Rate and Direction of Technical Change, 50 Year Anniversary, Edited by Josh Lerner and Scott Stern, University of Chicago Press. Pp 203-276.)

The news this morning announced Bill Lowe’s passing. I am greatly saddened to learn of his passing, and my sympathies go to his many friends and family. He had a unique role in computing history. In this post I would like to share a few memories of those interviews.

******************

Just Finish it

Walk into any running store and tell the employees you will run a marathon or half-marathon. Inevitably that employee, who is usually a runner too, will respond in a hushed tone with a measured statement, like they are imparting a truism, but afraid of offending your religious sensibility. They say, “I prefer the distance of a half-marathon. That is a good distance.”

I think I finally understand that statement, and why everyone says it with such reverence. It is a lesson that can only be learned the hard way. I learned it this morning running the Chicago Marathon.

I ran the Chicago Marathon last year. It was my first marathon. The bucket list figured prominently in the motivation for trying that race.

The bucket list does not explain why I ran the Wisconsin Marathon last spring, nor why I ran this morning. The motivation remained difficult to articulate. I will get to that momentarily. I mostly wanted to satisfy my curiosity.

Yes, most people do not train for and run marathons just to answer a question. That will take some explaining.

Today’s blog provides a discourse on the motivation for running marathons. Oh, it also has details about how today’s went.

Digital Public Goods

Precisely how does the online world provide public goods? That is the question for this column.

Public goods in the digital world contain some of the same features as those in the offline world. Yet, there are some key differences in the boundaries between public and private, and that shapes what arises and what does not.

That will need an explanation.

The economic policy of data caps

It is the one year anniversary of the Open Internet Advisory Committee (as noted earlier). Today the committee issued a report of its work over the last year. You can access it here. Today’s post discusses the report about data Caps, which was written by the Economic Impacts working group.

I am a member of the committee and the Economic Impacts Working Group, and I like the work we did. I chair the group. “Chair” is a misleading title for what I really do, which is take notes of the groups’ discussions and transcribe them. Every now and again, I do a little more. As one of the members without any stakes in the outcome, occasionally I offer a synthesis or compromise between distinct views.

The report aims to analyze data caps in the context of the Open Internet Report and Order. The Open Internet Report and Order discusses usage-based pricing (UBP), but does not expressly mention data caps except by implication in that data caps can be considered a form of UBP. The Order left open the possibility of many experiments in business models and pricing.

Moreover, the Internet had evolved over time, and the Order anticipated that the Internet would continue to evolve in unexpected ways. The Order set up the advisory group to consider whether aspects of the Order remain consistent in its effects on the Internet as the Internet evolves, and it is in that spirit that this conversation was undertaken.

The Open Internet Advisory Committee at year one

Today I would like to make a little shout-out for recent work at the FCC to improve policy making for the Internet. To do that I need to put my preferences front and center.

There are policy debates, and then there is actual policy making. The former grabs headlines on occasion, while the latter rarely does. Both need to take place in order to make progress, albeit, it is a rare person who has the patience and taste for both.

I have little patience for the grandstanding that goes with policy debates, and I do not take much pleasure from the staging and entertainment behind political posturing. I prefer policy making, especially the quieter and more challenging parts of it, and I love being engaged in challenging policy conversations that do not get much publicity.

Just so we are clear, this post will discuss policy making. Policy debate will largely remain in the background. That is unusual for most public discussions about policy for the open Internet, but it seems appropriate for today’s post.

It is the one year anniversary of the Open Internet Advisory Committee. In approximately two weeks the committee will release its first big report, a kind of year-in-review. I am not a neutral observer of this committee. I am a member. I am especially impressed by what the committee did in its first year.

If you think I am biased, then you are right. That is the point of this blog post.

I have been happy to be part of this committee, and contribute to public policy discussions through participation. And whatever else the posturing political world says, I want to be the first to say loudly that this committee has done wonderful work to support policy making, and, until two weeks from now, largely out of the public’s eye.

Information Technology in the Desert

Every summer my wife and I take the children west to experience nature at its grandest. Though these trips are designed to foster “quality family time,” invariably they teach us more about ourselves than merely about nature. In previous posts I have used these trips to learn something new about the role of information and communications technology in our lives.

This year we visited the southwest. The Hoover Dam served as appetizers then we turned to the main course –Zion, Bryce and the Grand Canyon. The Petrified Forest and the Painted Desert served as dessert. Yes, indeed, the desert was dessert. (Sure, it was clumsy, but worth the artlessness nonetheless. How many times in your life do you ever get to put those two words in one sentence?)

This year I learned two kinds of lessons. One was about connectivity, while the other was about prediction. That is not obvious, to be sure. Give me a minute. I will get there.

Differentiated Platforms

Differentiation is a standard concept for analyzing competition. It describes a common situation, where one firm develops the ability to serve one type of customer in a market—say, buyers who will pay a lot to save time—while a competing firm serves another—say, budget-conscious buyers who are patient.

Differentiation can describe common competitive behavior in technology markets. A chip firm might develop particular attributes—say, faster, energy-hungry electronics for a particular purpose—while their rival might specialize in slower chips that use little energy. This differentiation can earn each firm loyalty from buyers with different preferences.

That motivates today’s question: Can platforms differentiate? Platforms have played an increasingly important role in technology markets in the last decade—in mobile devices, in web services, you name it. A mix of standards composes a platform, complementing many other firms who build services upon the standards.

At first blush the answer appears to be yes. Think of attributes associated with common platforms, such as Windows, Android, Linux, Facebook, or the iPhone. These platforms differ from one another in the marketplace and set themselves apart from near rivals, in ways that earn the loyalty of particular users.

That first impression makes it worth a deeper look. There is more here than meets the eye, and smart firms shape their strategies with subtle thoughtfulness.