Back to article
The Yin and Yang of Open Source Commerce, Part 3
Chewing Over the Numbers
November 3, 2005
The last two articles dealt with the nature of the global IT market. These confirmed the success of Linux-based business so far, and yet the immense opportunity still open to those same businesses. This part explores that opportunity further.
A clear picture of Linux achievements can be obtained by comparing market share for operating systems over time. I have provided my own estimates that includes multiple Linux installation from a single media source, as well as IDC estimates for 2002, as presented by Sam Dochnevich, IBM, at the Desktop Linux Conference in Massachusetts in November 2003 (see Table 1 and Figure 1).
The growth in the Linux server market is greater than the losses of NetWare and UNIX servers. It is important to note that the Microsoft Windows Server installed base continues to grow, also. Surely the focus by the Linux companies (Red Hat and Novell) on the UNIX marketplace must be considered misplaced, because while Linux has gained ground in this market the question must be asked: "If the same effort had been expended at replacing Windows NT/200X servers, would Linux today have a larger share of the overall server operating system market?"
Jean Bozman, IDC, said that the overall server market grew in 2004 by 10.9% by value. Sales of Windows-based servers grew 14.3% by value and 10.9% by number. By comparison, Linux server sales grew 45% by value and unit shipments grew 32%. However, it is interesting to note from this report that UNIX server revenue grew 2%, while the number of units shipped slipped by 8.7%.
During the years 2003-2005 neither Red Hat or SUSE have specifically targeted the SMB/SME marketplace with an office infrastructure solution. Had they done so, the Linux platform might have acquired a simple to configure, powerful installation, and management system much earlier. Novell SUSE YaST2 has come a long way, but it is not even close to the Windows Server 2003 configuration wizard.
From the October 2005 Netcraft Web server survey, approximately 70% of the world's 74.4 million web servers use the OSS Apache web server. Approximately 50% of the Apache installed base runs on Linux. The calculated Linux web server installed based is thus 26.4 million web servers.
My personal estimate is that there are approximately 16-18 million Samba servers, of which in the vicinity of 11 million run on Linux. The vast majority of these Linux/Samba servers are in use in small businesses and in home networks. The enterprise Samba customer is more likely to use Sun Solaris, IBM AIX, or HP-UX. The SMB/SME market has deployed a significant number of Samba servers, however reliable statistics do not exist as to the number of Linux/Samba servers in use in this market.
From the statistics that are available, we may seriously question why the projected 2006 Linux server install base will be only 9.9 million units according to IDC. The fact that so many Apache web servers run on Linux proves one market in which Linux has become a dominant platform and thus a serious player. Companies that focused on web services found the world as their oyster-bed, if they could just find the right combination of software and services.
Had Linux companies better focused their efforts on the SMB/SME market, the operating system market share picture would look far more rosy for Linux and OSS than it does today. There is some suggestion in what we have seen so far that Linux has performed well despite the efforts of the Linux vendors. It is as if the market has made a run for Linux, in spite of the lack of market presence by Linux vendors. It would further appear that most of this has happened outside of the enterprise market space that was at the center of the quest to prove that Linux is enterprise ready--whatever that means!
Linux has become a most stable and compelling technical product and platform, as is evidenced by the exceptional rate of adoption. The rate of Linux adoption does not show any signs of slowing down. It is astounding that despite the fact that Linux is more difficult to deploy than a comparable Microsoft Windows product, clearly customers have hankered after it. It is a fact that there is a huge market opportunity in the SMB space that is begging for an alternative solution to the one they are using right now.
Being able to identify a market does not help if the right products and services are not available. Additionally, it does not help unless there is a cost effective method of delivering those products and services to the business consumer. We must look at the existing IT market to learn how current products and services are delivered to the customer.
There are four core IT markets:
It is informative to see what are some of the characteristic buying behaviors of each market. I am indebted to a selection of VARs, each of whom gave considerable time to provide feedback and comment for this series. A representative of each VAR was contacted and asked a to describe the predominant opinion in his office. The number of employees in the VARs' offices varied from 1 to 11. The findings from the first interview were collated anonymously, then each person was asked to comment on the feedback obtained from the others. The resulting comments are summarized in the next three sections.