A few weeks back, several leading executives in the high-tech community descended upon the Westin La Paloma resort in Arizona to attend the 20th annual PC Forum, a conference that focuses on the leading issues in PC land. As we observed (it was our first visit), we were struck by the ironic notion that no one seemed very interested in PCs. In fact, the agenda was focused more on education, privacy, and "communities" than transistors, software, or switches. And then there were media. No sooner do ads appear in cyber-space, than every high-tech executive becomes an authority on the media world.
It would not be fair to say that PC Forum had no representatives from the media industry. Two Michaels (Bloomberg and Crichton ) did attend. Not surprisingly, they both seemed overwhelmingly convinced that this group of high-tech leaders was several sandwiches short of a picnic when it came to understanding the underlying aspects of the media world. We are quite sympathetic to this attitude. However, while we recognize the potential for media naiveté, we also believe that the media world may underestimate the influence technology may have on that world. For instance, consider that the economics of running a new media Web site are much more like those of a software company than those of a traditional publisher. Variable costs are near zero, and as a result, market share means everything. Media leaders also lack an appreciation for the concepts of technological lock-in and platform leverage. In the technology world, the best product does not always win, and the same may hold true for new media.
We would like to begin our journey by dividing the new media content world into three categories--commodity information, editorial information, and "hits." Commodity information includes general information such as stock prices, sports scores, and news stories. Editorial information includes those sites we look to as definitive sources on particular genres, topics, or products. Lastly, hits represent new media's attempt to recreate the popularity of television shows.
Let's begin with commodity information. There is no questioning the value of having immediate access to commodity information. However, the value offered in this market is tilted more toward presentation and ease of use than ownership of the particular content. Stock prices, news stories, yellow pages, and door-to-door maps can all be acquired on a wholesale basis. Therefore, the market for these services is quite crowded. Also consider the following. First, we think this market is extremely susceptible to entry from Microsoft MSFT . Active Desktop will likely put these items directly on your computer's wallpaper. Second, this information fundamentally belongs on a broadcast network such as a satellite or cable network, and not on the Internet (for more on this topic see ATC 96-16).
Let us now address the Internet "hit" business. We are very skeptical that we will ever see "hits" on the Internet. Consider the fact that cable television represents a level of fragmentation beyond traditional television. How many hits are there on cable? Sure, there is the Larry Sanders Show, but this is a far cry from Seinfeld or The Simpsons. The Internet represents a level of fragmentation beyond cable, which is orders of magnitude beyond television. The Internet is not about mass audiences, and it's a good thing. The underlying architecture wasn't designed to support them.
This leaves editorial--the "sweet spot" of Internet content. Companies that are the leading providers of information about specific areas of interest will attract the majority of advertising dollars and transactional fees. Unlike commodity information, creation of this content requires work. And unlike "hits," this market is evolving nicely on the Internet. Editorial sites help people make better and faster decisions, and advertisers love to strike when the iron is hot.
We will now shift to Internet advertising terminology. The first term one must understand is page views. This represents the number of content pages that was actually delivered to customers. Page views are also referred to as "inventory," as they represent the potential number of ads that can be sold. These page views, or inventory, are sold to advertisers on a "cost-per-thousand basis," also known as CPM. Of course, multiplying CPMs by page views will likely generate a number far greater than that reported as revenue by the leading new media companies. This is because most ads are sold at a discount to the actual price, and because the entire inventory is hardly ever sold. Inventory utilization, or sell-out, represents the amount of page views that are actually converted into revenue.
Most of the leading new media companies have sell-out ratios well below 1.0, and many of the leading search engines have rates below 20%. An Internet advertising optimist would proclaim that this is a good thing--that unsold inventory represents opportunity, and that sites that are "sold-out" have limited potential. We call foul on this conjecture. For starters, sold-out sites can always raise rates. More importantly, it seems somewhat obvious that a product in 5-1 oversupply is highly susceptible to price erosion.
Some have suggested that America Online's AOL entry into the advertising market will be a catalyst that affects all CPM pricing. This is not necessarily the case. We doubt that if The Daily News lowered pricing, Fortune would feel much of an impact. The more likely driver of CPM rationalization is simple economics. Page-view inventory is transient, and cannot be sold next quarter. Therefore, it only makes sense to maximize the revenue opportunity by selling ads at lower prices. Markets should clear where supply equals demand. The interesting thing is that a price reduction could actually aid revenues. At a 20% sell-out rate, it is easy to imagine a scenario where prices are cut in half, but revenues increase.
One popular argument in the new media field is whether ads should be sold on a CPM basis or on a "per-click" basis. We would like to take a definitive stance on this issue and declare "both." Sites that are close to inventory sell-out will undoubtedly have price control, and therefore can sell on a per-impression basis. However, sites that are not sold-out should be looking to alternative methods to maximize revenue. Have you ever noticed that you see a large number of "1-800" ads on late night television, as well as on CNBC? The reason for this is that most of these ads are sold on a per-fulfillment basis. The content companies decide which vendors typically generate the most sales and fill their surplus air time with the most productive ads. It is with this in mind that we predict "per-click" and "per-sale" ads will proliferate broadly on the fragmented Web. Think of it as a type of advertising auction that will help determine the proper clearing price.
We will close with our evaluation of the search engine companies. There are several interesting things to note about search engine companies. First, they originated as editorial content providers, with Internet sites as their area of focus. Over time, they have added other areas of editorial focus, as well as their own presentation of commodity information. There are two more things that typify search engines. First, most of these companies have extremely low sell-out ratios, hovering at or below 20%. Second, page views are growing much faster than revenues. This indicates that either pricing (CPMs) or sell-out ratios are falling.
The core competency of the search engine is in providing editorial information about Web sites. This task is highly valuable, and while we have no definitive proof, we speculate that the majority of the search engine revenues comes from selling ads that are associated with particular search terms known as keywords. This type of business if very similar to that of a yellow pages company. Advertisers pay for predominantly displayed ads that are located specifically around certain keywords. And although we fundamentally believe that keyword ad rates are likely to rise, most search engine companies shy away from the "yellow pages" analogy (most of them refuse to disclose the percentage of revenues associated with keyword sales). This is despite the fact that the traditional yellow pages business is both quite large and quite profitable. We can only speculate that these companies find the "media company" label more endearing than that of a yellow pages company. RBOC's sell yellow pages, companies like Disney DIS sell media.
We offer one last thought regarding Microsoft and the potential ease of entry into the search engine space. Recognize that the variable cost of a copy of Internet site guide information is virtually zero. With companies such as Excite XCIT increasingly interested in new forms of funding, is it all that unreasonable to speculate that Excite could sell an entire copy of its site guide to Microsoft for, say something like $10 million dollars cash? Then, simply apply the leverage of desktop ownership and we would have a very competitive new player in the market.
J. William Gurley 1997-8. All rights reserved. The information contained herein has been obtained from sources believed to be reliable but is not necessarily complete, and its accuracy cannot be guaranteed. Any opinions expressed herein are subject to change without notice. The author is a general partner of Hummer Winblad Venture Partners (HWVP). HWVP and its affiliated companies and/or individuals may, from time to time, have positions in the securities discussed herein. Above the Crowd is a monthly feature focusing on the evolution and economics of high technology business and strategy.