WebMethods in WebMadness: more arithmetic in a bubble

"Arguing in My Spare Time" No. 3.06

Arnold Kling

Feb. 18, 2000

May not be reproduced commercially without permission of the author

WebMethods is a stock that went public earlier this month. I immediately entered WebMethods into the Internet Bubble Monitor calculator, replacing my sentimental favorite, MusicMaker, whose price from now on cannot fall by more than $5, since that is where it currently is trading.

WebMethods is riding on two hot buzzwords, B2B and infrastructure, as well as a potent sub-buzzword, XML. XML stands for "extensible markup language."

XML is designed to make it possible to write descriptions of data fields that can be interpreted automatically. This in turn potentially provides solutions to problems as disparate as (a) creating interfaces between two companies' legacy systems and (b) enabling sites to be described much more precisely to search engines, which would improve the accuracy of the latter and increase the visibility of sites with the best content.

I am bullish on XML, and in fact I have written that in five years we may look back on XML as more important than Java. I also agree that business-to-business commerce (B2B) represents a more solid opportunity than business-to-consumer. Finally, from the very first of my essays, I have argued that businesses are going to shift from proprietary systems development to generic solutions using Internet standards, which means that I buy into the "infrastructure" buzzword as well.

So I have a great deal of respect for what WebMethods is doing. The place they are planting their flag has solid ground underneath, as opposed to an illusory fad such as Data Mining or Open Source Software.

But the WebMethods business model is missing something. What is lacking is a story of growth with near-zero marginal cost.

For example, Netscape had the potential to enjoy effortless growth. Suppose that they had captured a leadership position in the market for Web server software. As demand for this software expanded, the only impact on Netscape would have been to require the company to keep enough ftp servers running to ship the product. Instead they could not compete with Apache (yes, I know that represents a win for "open source.").

Three years ago, Yahoo was in an effortless growth position. Web usage in the United States was increasing exponentially, and for Yahoo each new user added positive marginal revenue at zero marginal cost.

WebMethods is not in that game. Web Methods sells "solutions" to large enterprises. This means a combination of software and consulting. Each new customer requires additional labor, both for sales and to provide services.

The most comparable company to WebMethods might be Computer Associates, a computer service firm that also offers a combination of products and consulting services to big companies. Computer Associates' historical strength was in the mainframe industry, but through acquisitions it has built capabilities in all facets of computer systems.

On February 17, when WebMethods stock hit 300, giving it a market capitalization of $9 billion, Computer Associates had a market cap of $37 billion. Computer Associates had sales in its latest fiscal year of $5.3 billion, giving it a price-to-sales ratio of 7. This probably represents a reasonable view of profitability in their business. Computer Associates has a healthy price-earnings ratio of slightly over 50, so if its price-to-sales ratio is low it is not because the stock is in the tank.

In a steady state, what will be the ratio of WebMethods' market cap to sales? I would think that it is likely to be closer to Computer Associates' ratio of 7 than to 2000, which is where WebMethods' ratio stood when its price was at 300, based on 1999 sales of $4.5 million.

To simplify the arithmetic, let us assume a price-to-sales ratio of 9 and a market cap of $9 billion. That means that WebMethods needs $1 billion in sales to justify its market cap.

Another ratio that might be relevant to WebMethods is Computer Associates' ratio of sales to employees. With 18,000 on staff, that ratio was roughly $300,000 per employee.

Assuming that WebMethods also reaches a steady state ratio of $300,000 in sales per employee, the company will need over 3,300 employees to achieve $1 billion in sales. When it went public last week, WebMethods had 150 employees.

The fundamental issue for WebMethods, assuming that they do not suffer from "core incompetency" as Netscape did and actually can deliver what the market is looking for, is how they get from here to there. It seems obvious that in order to grow the company by a multiple of twenty, they are going to have to get very good at one or both of the following:

• acquiring companies at prices that are cheap, or at least no more than fair

• attracting, training, and retaining a high-quality work force.

In fact, it is competence in these areas that has made Computer Associates what it is today. I am not saying that WebMethods does not have those skills. But they have not been tested. Even with all the right buzzwords going for it, a 150-person firm in the computer services field with a market cap in the billions is just more air inside a bubble.