On April 16, 2008, The Standish Group released a report titled “Trends in Open Source.” That same day, we put out a press release that said, “Free Open Source Software Is Costing Vendors $60 Billion.” There has been much blogging and discussion around how we came to this conclusion.
The report itself discusses The Standish Group’s research study of the top 10 drivers that are influencing decisions on how IT is adopting open source technology, and very little on the size of the open source market. In fact, the only place that the report addresses this area is in the summary, with these few words: “If product and service revenue were calculated at commercial prices, the Open Source Initiative would be the largest software company in the world, outselling Microsoft, Oracle, and Computer Associates combined.” In 2007, Software Magazine’s Software 500 ranking showed the combined software and services revenue of Microsoft, Oracle, and Computer Associates to be $58 billion.
The Standish Group relies only on its own original qualified research and uses no secondary research. The main instrument for this report was our Demand Assessment Requirements Tracking Survey (DARTS) on Open Source, conducted during the month of February for the last five years. This article provides a view into our thought process and how we came to our estimate of $60 billion.
Method 1 (Ongoing): Worldwide IT spending is more than $3 trillion per year, and, before this latest slowdown, had been growing at around 6 percent per year. Our DARTS numbers indicate that CIOs estimate that they save 6.2 percent of their IT budget based on using open source products. (See Fig. 1.) If we simply take the $3 trillion and multiply it by 6.2 percent, we get an average savings of $186 billion. This gives us our first savings range of $186 billion.
However, we know that ongoing cost equals 70 percent of the IT budget. This gives us a base number of $2.1 trillion. If we take our 6.2 percent savings, this would lead us to conclude that the ongoing savings is $130 billion. Later we need to add back the 30 percent for new projects. A third of the $3 trillion of the IT budget is product spending. Taking our 6.2 savings percent and multiplying it by $1 trillion gives us $62 billion in product and service savings. This money would normally go to product and service vendors. Therefore, using this method, the range is between $62 billion and $186 billion.
Method 2 (Product List): First we listed the major open source products. Then we looked at the commercial equivalents. Next we looked at the average cost of both the open source products and the commercial products, giving us a net commercial cost. We then multiplied the net cost of the commercial product by our open source shipping estimates. This gave us a number below the $60 billion, but we then considered other open source products and services. Plus we needed to look back at previous years to consider the installed base for the loss of add-ons, upgrades, and service revenue.
When we do these additional calculations, we rapidly get more than $60 billion in lost revenue. In addition, price pressure is being put on all commercial software. This is good for the user, but bad for the supplier. These reductions should also be considered as free and open source software (FOSS) savings, even when you are not using FOSS.
Method 3 (Market): We then took another look at the major open source products and the total market for that segment. For example, Linux is 21 percent of the operating system market. First you need to recalculate the market based on adding the 21 percent back and then taking 21 percent of that market. This again gave us a number below the $60 billion, but adding upgrades and service to the base gave us a number larger than $60 million.
In addition, we needed to consider the price reduction because of the open source alternative. For example, Microsoft Office was selling for $400 per copy, but now you can buy a home version with three copies for $120. I bought this software, installed it on my Mac, and I am currently using it to write this article. I also used the other two copies. This normally would have cost me or you $1,200—a factor of 10. I doubt this reduction would have happened without the pressure from Open Office and Star Office.
Method 4 (New Projects): From DARTS, we learned that our CIOs believe they save 11 percent of new requirements through the use of open source components. The software service market to build new software is around half a trillion dollars. This would bring the savings on loss of services to $55 billion. Our CIOs say that they are saving, on average, 5 percent of their new project development costs.
If we go back to our $3 trillion in spending and look at the 30 percent for new projects, we can conclude the yearly savings to be $45 billion. This number should be added to our $130 billion for a total savings of $175 billion. This gives us a nice range of savings from $175 billion to $186 billion. A third of this number for product savings seems reasonable, giving us a range of $58 billion to $62 billion.
Method 5 (FOSS Effect): The FOSS effect is the ratio between the cost of the open source product and the commercial product. (See Table 1.) For example, one of our clients moved part of its operation to a full stack of FOSS. The operation had been running on a closed stack, and the software amortization, upgrades, and maintenance costs came to $7.5 million per year. Now, that company’s support costs and other fees for FOSS are half a million dollars. How would you account for that transaction? Do you only look at the half a million dollars as a new sale? That would be the increase in the FOSS market, but it is a $7.5 million loss to the closed software market. That revenue is gone, and gone every year—forever.
In this one example, the ratio is 15-to-1. This might be an extreme number, but there is a correct number, and for every $1 in FOSS, you lose X dollars in closed software. Let’s say the X is equal to 10 and the market for FOSS software and services is $3 billion. Then the loss to closed software is $30 billion.
However, you need to consider the installed software from previous years. In going back five years with modest growth, this would equal more than another $30 billion in add-ons, upgrades, and maintenance.
These were the five methods we used to estimate the true size of the open source market as it compares with the closed source market. We believe this number to be $60 billion. However, as our lawyers suggest, all data and information in this article should be considered Standish opinion, and the reader bears all risk in the use of this opinion.
It should also be noted that DARTS respondents are CIOs of mostly large organizations. First, they tend to be conservative on their estimates. Second, 38 percent reported no open source products. Zeros have dramatic effect on lowering the average. However, we know from our workshops that while CIOs say there is no open source software, there is plenty installed that they do not know exists. Therefore, again, we believe our number of $60 billion to be conservative. The savings to the overall IT budget is much, much greater.
As we said in our “Trends in Open Source” research paper: “Open source technology is providing organizations with the ability to rapidly implement new business solutions. Organizations throughout the world are saving billions of dollars per year by using and implementing open source solutions. Other organizations are earning new revenue and satisfying users’ expectations through the use of open source solutions. The open source movement is advancing because of feature-rich, secure, high-quality, reliable software with compelling economic benefits.”
For a copy of the “Trends in Open Source” paper, go to www.standishgroup.com and click on the market research link or the store link.
Jim Johnson is the founder and chairman of The Standish Group. He has been involved in the computer industry for more than 30 years and has a long list of published papers, articles and speeches. Go to www.standishgroup.com.
