Wednesday, September 27, 2006

Google video lectures - not yet properly indexed!

In collaboration with Google, the University of California at Berkeley is making available lectures on a range of topics at

This is a very interesting development and there are a number of talks relating to the Internet, mainly to do with issues about search. Unfortunately several of the talks seem to have been incorrectly indexed by Google!

I have attempted to correctly align the links below.

Search advertising Hal Varian
Intellectual property and search Jason Schultz
Search engines, technology and business Sergey Brin (cofounder of Google)
User experience issues in web search Daniel Rose
Quality and search Geoff Nunberg
Web spam Marc Najork

Thursday, September 21, 2006

Critical mass

One of the terms that you come across in discussions about the growth of the Internet, or of Internet applications or services (for example the use of Internet telephony or the adoption of a new browser like Firefox), is critical mass. Once this critical mass has been reached there are enough customers in the market using the product or application for the positive feedback effects associated with network externalities to really kick in and for the product to takeoff. Until this critical mass has been reached there might still be doubts about whether the new product will establish itself and so potential new users may be reluctant to get on the bandwagon.

There are plenty of examples of products where this critical mass was never reached and so the product eventually failed and was withdrawn from the market altogether. An example described by Rolhfs in his book
ROHLFS, J (2001) Bandwagon effects in high technology industries is Picturephones in the 1970s. As he explains, it was technically possible to make telephones that displayed a picture of the caller on a screen, but the product never took off - critical mass was not reached. (Of course today mobile phones that take pictures are all the rage but that is a completely different type of product).

There are a number of reasons why critical mass might not be reached for a product. It might be too expensive in relation to the features that it offers for it to be attractive to buyers. Or it might have been poorly marketed. That is why economists often advise firms with new products that might exhibit network effects to keep the price low early on even if that means that costs are not covered, and to spend money on marketing and promotional campaigns to get the product talked about. So long as the product will be profitable once sales have passed a certain level then it will be worthwhile in the end. The price discounts and marketing costs can be viewed as an investment. You might even see a product given away for free in the early stages of its life as a way of pushing up its use beyond the critical mass. Once this level of use has been achieved other people will want to have it too and you can start to charge for it. There are lots of examples of software and web-based services that were given away free in this way early on but then require payment or subscription fees once it had become established.

How big is the critical mass?
An obvious follow up question that might be asked at this point is "how big is critical mass?" - either in terms of the number of users or the share of a market. There is no easy answer to this question. It depends on the size of the market and its competitiveness.
Here are some examples from recent news reports.

  • Firefox at Critical Mass? Jim Wagner, Internet News, 6th January 2006.
    Wagner reported that Firefox had achieved a market share of 9.57% in December 2005, just under the 10% figure. He quoted Vince Vizzaccaro, executive vice president of marketing and strategic relationships at NetApplications, the company that collected the data "Firefox is very close to hitting a critical mass of 10%, which could mean a more rapid adoption rate".
  • US Broadband Market Reaches Critical Mass Kirsten Fischer, In-Stat Press Releases, 6th April 2004. The report says that "with close to 27 million US business and residential subscribers, broadband is now clearly a mainstream service". It goes on to mention some of the applications such as home entertainment, VoIP and online gaming that require a broadband connection and quotes Daryl Schoolarm a Senior Analyst at In-Stat "This starts a cycle where growth in both broadband and applications feed the growth of each other". This is what Rohlfs means by complementary bandwagon effects.
  • MP3 Players Reaching 'Critical Mass InternetWeek 13th April 2005. The story quotes JupiterResearch analyst David Card as saying that the number of MP3 players (in the US) was close to reaching around 18 million. "Historically, any new device or medium that reaches a US household penetration of 15 percent to 20 percent creates a critical mass of customers for other products and services. This is good news for digital stores and subscription music services. Subscription services and devices will fule each other's growth."

Some references to critical mass seem to imply that it is the level at which a product becomes profitable. This is not quite correct. In an established market there will be a break-even point beyond which revenues exceed costs and so the firm sells enough of its product to make a profit. This is related to economies of scale. Enough units are being sold for the fixed costs of production to be shared over many units so that average costs come down. In a new market with an untried technology the cost of production might initially be high in any case. Costs can come down as the market is established partly due to economies of scale, but also because the production process itself is likely to be improved once the firm moves beyond the pilot stage. What is correct then is that a firm will probably not be able to earn profits unless it can reach a critical mass.

Another myth is that having a big market share is necessarily going to lead to profits. As John Kay noted in his article in the Financial Times back in March 1998, the share of large firms in total output is not going up. He underlines his point by identifying the automobile industry as one which has become less concentrated over the years despite it being an industry that is often used to illustrate the importance of economies of scale, market concentration and critical mass.

Footnote: the term critical mass itself probably comes from nuclear physics. It identifies the amount of plutonium needed to start a chain reaction.

Monday, September 11, 2006

Different dimensions of the digital divide

According to some recently released figures almost 60% of households in Great Britain (13.9 million) now have home Internet access (source: Internet access: households and individuals, Office for National Statistics, 23rd August 2006). This represents a big jump of around 26% since 2002 and is around 5% up on last year's figure. Moreover nearly 70% of households with Internet access have a broadband connection. To put it another way about 40% of UK households now have broadband Internet access.

It would appear that the government strategy of leaving it to the market to deal with the matter of Internet access, with government limiting its role to providing appropriate regulation via Ofcom to ensure competition, has been working. However these national figures do hide some noticeable regional variations. For example, while two-thirds of households in London and the South-East of England have some kind of Internet access at home, in Scotland and Northern Ireland the figure is closer to 50%.

The reasons for this geographical digital divide are varied. The problem of reach has now largely been overcome. At one time there was a marked urban/rural divide and this to a large part could be explained by the inability of the Internet Service Providers to provide adequate connectivity to certain places in the country. Although there are a few remote locations where Internet access is physically impossible, some kind of link either via telephone lines, cable or even a wireless network is now available to most of us. So the remaining differences must be explained by non-technical factors.

One significant factor is the age of the members of the household. The ONS report reveals that 83% of people in the age range 16-24 had accessed the Internet in the three months prior to the survey, but only 15% of those in the 65 years and over group had done so. There are gender differences too - 40% of women surveyed had never used the Internet while only 30% of men had not done so. Respondents were invited to give reasons why they had chosen not to arrange home Internet access. About a quarter of the sample stated that the Internet had no interest or value for them - and about the same proportion said they lacked the skills. It could be that these are the same people, lacking both an awareness of what can be done on the Internet and the knowledge of how to use it.

Other respondents replied that equipment and/or access costs were too high. The price of PCs and laptops has fallen substantially in recent years, and the price of Internet access is also coming down as the market expands and becomes more competitive, so there should be further scope for the proportion of households with Internet access to rise still further in the future. However household income is clearly important here. A few hundred pounds may be easily found for a PC in an affluent household but impossible to justify in a poorer one. Class and race may also be a factor. Just today I came across a report on a study in the US that found white children are far more likely to use the Internet than are black and Hispanic children. (See Racial digital divide lingers among US schoolchildren, Jennifer LeClaire, TechNewsWorld 11th September 2006).

You can find many other interesting contributions to the debate about the digital divide on the web. Here I highlight just a few. On the Ofcome website you can find a Report following the Remote and Rural Communications Symposium in February 2005. There is also plenty on the Ofcom website about the need for greater liberalisation in the telecoms market to ensure that there is an appropriate market structure to guarantee efficiency in the market.

The ISOC website links to a contribtion by Madanmohan Rao
Struggling with the Digital Divide: Internet Infrastructure, Content, and Culture. Is a progressive Internet environment enough to close the gap between North and South?

The Pew Internet and American Life Project provides evidence on the age dimension to the digital divide in the US - see for example Wired Seniors: A fervent few, inspired by family ties, September 2001.

Finally go to the UN's website and read what Kofi Annan wrote back in November 2002 On the digital divide.

Thursday, September 07, 2006

Internet developments and enduring issues

This week I welcome another new group of students to one of my Economics and the Internet classes. Since September 1999 I have been teaching final year undergraduate students on the BSc Business Economics at the University of Surrey on a module called The Internet for Business Economists. I also teach a similar module called Economics of the Internet to postgraduate students at the University of Portsmouth. As the Surrey course comes in the first semester and the Portsmouth one comes in the second semester I have to update my material twice each year, which is just as well as Internet related matters change on almost a daily basis. Brad DeLong once said (1998) that "Competing on the Internet is like leading a dog's life - it compresses seven years into every one". Things can change very quickly and there is a lot to keep track of.

The other courses that I teach also need updating each year, of course, but the changes that are needed are usually very minor. I might have to add in a couple of references to some recently published articles or a new edition of the textbook but this can usually be done at a single sitting. But with the economics and the Internet courses I might even have to change stuff in the week that I give a lecture to recognise a new advance, to acknowledge a new milestone in the development of the Internet or to incorporate the latest figures on the amount of e-commerce. If you want to keep up to date with developments it is a good idea to sign up to one or more news alert service, which comes either in the form of a regular e-mail newsletter or as a news feed such as the ones from the BBC, Google, The Register or E-Commerce Times. You can specify appropriate keywords such as Internet, e-commerce etc. if you want to narrow down what you get. Some other sources that I find it very useful to watch are the Pew Internet and American Life project, Nielsen-NetRatings, and Information Week. For links to up to date statistics see my earlier blog Internet statistics. Of course I will try to use this blog to direct you to important news that is relevant to the course.

But even though the technology may change there are plenty of enduring questions that remain the same that we shall be considering on the course. How can economics help us understand the rapid growth of technologies like the Internet and e-commerce? [See my blog on bandwagons and network effects]. Would an e-stamp help in the fight against spam? [See my blog Controlling spam via e-mail charges]. What has been the impact of the Internet on the prices of goods and services? [I have got a blog on this too.] Why not browse now through this and all my other previous blog postings to get a better idea about what economics has to say on these and other Internet issues?