Dan Bricklin's Web Site: www.bricklin.com
Small Players Matter
While large players and big media companies act like they are the main reason for the web and Internet and therefore should drive policy decisions, the numbers show that the contributions of the myriad of small players -- individuals, non-profits, and small businesses -- are crucial to the vitality of the web and its value to people.
As part of my work at Trellix Corporation, I do a lot of thinking about small businesses (such as my Small Business and Web Sites article). As part of being a commentator on the general Internet, I've discussed volunteer labor (The Cornucopia of the Commons: How to get volunteer labor), as well as the success of search engines like Google (What's been successful in B2C). Recently, discussions by David Reed on SATN.org and elsewhere have talked more and more about "option values" and opening up infrastructure to lots of companies as opposed to a few large players. This essay ties some of all this together.
My thesis is that the web sites of small players have been, and will continue to be, crucial to the growth and acceptance of the Internet in everyday life. These small players include small businesses, passionate individuals, and non-profits.
There seems to be a conventional wisdom that the Web is dominated by a few very large players and that other tiny players don't matter. Mass marketed "content" like Britney Spears and Star Wars is what really matters, not the amateur writings of "non-professionals". According to Nielsen/Netratings in April of 2002 AOL/Time Warner web sites had almost 65 million unique visitors spending 39 minutes each, while Microsoft web sites had over 63 million visitors for over an hour, and Yahoo! had over 59 million for almost an hour and a half. How can a web site like mine, with around 30,000 visitors a month, or a lawyer with 10, expect to "compete"? Aren't we the messy underbrush ready to be cleared out?
Let's look closer at the numbers. The Nielsen/Netratings usage numbers show that the average Web-surfing household in the USA had 21 sessions in April 2002, visiting 46 unique web sites, and spent over 11 hours doing it. Copying the Nielsen numbers into a spreadsheet, and multiplying the "reach" percent times the average time, etc., you find that the average surfer spends less than 2.5 hours on the top 10 web properties, less than 2.9 hours on the top 25. What do they do with their other 8+ hours? And anyway, are they reading the news, horoscopes, and other "professional" content, or are they using Yahoo! (#3) Mail and Microsoft (#2) Hotmail and therefore reading personal content? Are they looking at Yahoo!'s GeoCities or Lycos' (#4) Tripod personal and small business web sites? Is the very popular eBay (#8), with the largest average time per person of almost 2 hours a month, a big professional web site or a conglomeration of many small business web sites?
Much of the use of AOL (#1), MSN (#2), Yahoo! (#3), Lycos (#4), and Google (#5) is searching. Looking at search engine queries can give us some hints as to what people are doing on the web. During a conversation I had with Google CEO Eric Schmidt, he stated, as I recall, that most search queries weren't "common". In other words, while there may be lots of "Britney Spears", "sex", and similar requests, they make up a small percentage of the total, and most everything else is evenly distributed all over the map. People have very individual needs on the Web. Google has indexed over 2 billion pages, and gets about 3-5 billion search queries a month. There are probably over 40 million web sites (Netcraft found 37 million based on unique hostnames, and you need to add millions more personal and small business web sites that are sub-domains of places like GeoCities, Tripod, and Lawyers.com). Somehow, many of those web sites get read, too.
You can use the databases of Overture.com to see how many search queries were made in a recent month through the web sites they serve (MSN, Yahoo, and others). Overture sells "paid placement" search results. They let you see how popular various search terms are so you can plan your advertising campaign. They report that the heavily promoted Britney Spears (2,600K queries in April) is part of queries only 100 or so times more than the general term "Toronto Hotel" (23K) and just over 5 times more than "Holiday Inn" (491K). The term "hotel" was part of 1,600K queries, dwarfing Jennifer Lopez's 490K or Star Wars' 430K (both on the Lycos 50 list). The term "Bed and Breakfast" at 145K is more popular than Jennifer "Rachel" Aniston of the hit TV show "Friends" or her movie-star husband Brad Pitt (136K and 131K, respectively). "Lawyer" (117K) is just a little less popular than "teen sex" (200K). Even "tupperware" was searched for over 43K times. Simple terms like "accountant" (24K), "locksmith" (13K), and "waffle" (5K) add up when compared to the far fewer "popular culture" terms. Let's look at flowers (771K) and a few related terms like orchid (78.5K), azalea (39K), dogwood (15K), rose bush (14K), rhododendron (9K), tulip (5K), marigold (3.5K), and crocus (3K) -- it totals over 900K. To see even more variety, look at Ask.com's "Peek through the keyhole" page which lists the questions people are asking, updated every 30 seconds. Out of hundreds of millions or billions of queries processed by Overture each month, even "sex" seems unpopular since it's part of only 3.4 million requests. Of course, even the "popular" terms are often used to find web pages created by regular people and small businesses. If Disney or Yahoo! had all you wanted, why would you need Google?
The conclusion? Much of what we search for, and therefore use the web for, is not from large, dominant players. We read content from regular companies, from people with different passions (be it about a movie star, or a type of old car and how to get spare parts, or just about the events of their day), and from non-profit institutions (like hospitals, colleges, and universities) that put up detailed information.
The reason Google is so popular is that it works so well at finding what you want. The way in which it does that is look to content on the web for guidance. To rank the answers to your queries Google weighs links from other web sites, the text around those links, the position of those links on pages, and more. It uses computing power to tap into all the work people have done when they created their own web sites. While many "commercial" web sites like those from Disney, Amazon.com, and AOL mainly link to themselves or paid affiliates, many small business web sites and most personal and academic web sites link to other unaffiliated web sites that the authors truly feel are worth looking at for more information. That "volunteer labor" is what makes the web work (as eloquently described by David Weinberger in his new book) and makes Google possible. Google, in turn, showed that the "do it with people we hire and control" method of Yahoo! and others didn't scale to handle a growing web as well as Google's "distributed, web based, trust others" approach. Only by using the wide variety of the real people who make up much of the web can you find your way around that ever-so-rich real world. And, as the search query and other data shows, real people want that wide variety. I know what McDonald's is serving -- I need the web to let me know what that new restaurant in my town has as a special this month. And as to the volume: Remember that lawyer with 10 visitors a month? Visitors that typed in a search query like "divorce lawyer newton" are looking for something they need now. If the lawyer gets one new client every few months at $5,000-$10,000 each, the web site has more than done its job. (Many of the "top" web sites with millions of visitors are still struggling to make money on the web.)
While large players and big media companies act like they are the main reason for the web and Internet and therefore should drive policy decisions, in actuality they are just "the biggest of the many small players" that make up the Internet. In fact, the controlling "stay within us" mentality some of them have is actually counter to the needs of the Internet for growth. The numbers show that the contributions of the myriad of small players -- individuals, non-profits, and small businesses -- are crucial to the vitality of the web and its value to people.
- Dan Bricklin, 5 June 2002
Some follow up:
An analogy I thought of this morning: Treating big players like they matter so much more than the mass of smaller players is like looking at traffic on various roadways and deciding that only Interstate highways matter because there are few local roads that get anywhere near that amount of traffic. How can Main Street compete with the mighty I-95? Well, there are lots of Main Streets and that's where I get my clothes cleaned, buy my groceries, visit my nearby relatives, and use to get to the local ballpark...
- Dan Bricklin, 6 June 2002
Comments and reactions from others:
Google and Big Media's New Clothes in LawMeme by Ernest Miller. [6Jun02]
Economic Geography shapes the Web: Phil Wolff gets inspired to write more about search and small players: "The first promise of the web was to collapse distance. And it does. But we still commute. It's called search. You may enjoy searching. But search is usually an attention tax, as bad as pop-up ads between web pages. One of the reasons Google is so popular is that it takes less effort, time, confusion, and attention to get me to my destination." [7Jun02]
Private letter from Jakob Nielsen (with permission): "Since 1997, I seen much evidence to suggest that many aspects of the Web follow a Zipf distribution. This means that the biggest sites *do* get more traffic, but since there are so many more smaller sites, they add up to more when taken together. In particular, if we assume that the value of a pageview is higher for more targeted information, then the combined value of 10 million small-business sites is $1 trillion, whereas the combined value of the 10,000 biggest sites is only $19 billion (because they only generate 1 cent per page view because of their generic nature - in fact, after I wrote my analysis in 1997, the value of a generic pageview has probably dropped below a cent, so the value of the big sites may be even smaller than I calculated). See: Do Websites Have Increasing Returns? and Zipf Curves and Website Popularity." [7Jun02]
[DanB: It's pretty amazing to read 5 years later what he wrote about what he thought might happen in the year 2000 and comparing it to numbers today.]
Beyond the Portal: An old article by Mark Bernstein from the days when "...the popular wisdom said portals were the only Internet properties of value." Mark points to it in reaction to reading this essay. [11Jun02]
© Copyright 1999-2010 by Daniel Bricklin
All Rights Reserved.