We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.


Open Source, Outsourcing and Good Old-Fashioned Opportunity

Originally published November 9, 2006

Well before the dot-com boom and bust, the service bureau business model offered entrepreneurs a path to riches that actually worked:

  • Get investors and/or financing to buy mainframes that small/mid-sized businesses (SMBs) couldn't afford themselves.
  • Sell time-shared access to the hardware so each customer gets a fraction of the use of the expensive systems and gets the benefits of using those systems, without paying the huge up-front costs to buy them.
  • Provide value-added services to customers through access to software and applications, and leverage the investment in staff and applications by providing services to many different customers.
  • Profit.

Big profits were possible buying mainframes and selling time-share services. The service bureau owners did all the technical work and, in effect, hired themselves out as providing an MIS department at a fraction of the investment their customers would have had to pay to do it themselves.

Unfortunately for the service bureau operators, the entry cost barriers for SMBs has been dropping like stones for 50 years. These days, you can buy the processing power of a 1960's million dollar mainframe for pocket change, and fit the hardware in your pocket. There's no need to buy time-shared access to hardware when you can get it yourself for next to nothing.

But the other major component of the service bureau product is service. “Service” means not just access to the hardware but also access to services such as programming, custom software, user support and system administration. As profit margins for selling hardware have dropped, those vendors started to make up the shortfalls by selling more software and services, but now that no- or low-cost open source software is putting pressure on software vendors to reduce their licensing costs, vendors will have to make up the shortfalls in revenue somewhere else – by providing services.

Total Cost Allocation of Computing Services
Just because Gartner Group introduced the concept of “Total Cost of Ownership” (TCO) in 1987 doesn't mean the logic was new. Businesspeople down through the ages have made decisions based on overall cost-effectiveness rather than on a single dimension of costs, such as purchase price: tungsten steel drill-bits might cost twice as much as ordinary steel bits, but if they last five times as long as the ordinary bits, a mine manager would be foolish not to buy the more expensive bits.

What Gartner accomplished with TCO was to give IT managers a tool to build their case to choose the tools that would have the greatest positive impact on the corporate bottom line, rather than simply opting for the lowest bidder. Microsoft promoted its OS products as offering a lower total cost of ownership than their competitors at a time when Microsoft's competitors were other commercial software vendors like Novell (NetWare), Ashton-Tate (dBase II, III, IV), Lotus (1-2-3) and others, rather than providers of no-cost open source software such as Red Hat or Novell (SUSE Linux).

Now that Microsoft does face competitors who give away software, the TCO analysis becomes much more important. The Microsoft report “Get the Facts: Total Cost of Ownership” summarizes TCO issues raised by the choice of Linux or Windows. In the report, TCO is defined as including “...not only hardware and software acquisition costs, but also the costs of downtime, training, outsourcing, and ongoing system administration and management, with staffing costs being the largest cost component. TCO typically is calculated over the useful life of a system, with broad studies often using 3- or 5-year time horizons.”

In this report, Microsoft cites research from IDC (2002, “Windows 2000 versus Linux in Enterprise Computing”) showing that “...hardware and software costs combined represent only 9% of total costs, and that 86% of TCO is attributed to staffing and downtime – a dynamic that should lead organizations that want to minimize TCO to focus on ways to make IT staff more productive and systems more reliable.”

In other words, the lion's share of TCO is spent on the people who have to make the software and hardware do what they're supposed to and not do what they're not supposed to. Now that full-grown MIS installations no longer require millions of dollars investment in hardware, but they do require huge investment in people with skills, the targets for outsourcing have been reversed. If you could save 50% on software costs (according to this research) you'd only reduce your TCO by about 2%.

When viewed in terms of TCO, the cost of the software is almost negligible; certainly, reducing staff and downtime costs yields the greatest benefits, while spending time trying to reduce software costs becomes virtually irrelevant.

It just doesn't matter what software you use, because according to these results: a 10% reduction in software licensing costs reduces TCO by less than 1%, a 10% reduction in downtime translates into a 2 to 3% reduction in TCO and a 10% saving on staffing will reduce TCO by a relatively whopping 6%.

We're advanced sufficiently that software and hardware, formerly the big cost centers of MIS and IT, have been abstracted out of the picture: together, they compose less than one-tenth of the TCO of a software infrastructure. Downtime is almost a quarter of TCO, but downtime is largely a product of the work of the biggest component: people with skills – almost two-thirds of TCO.

Problem or Opportunity?
Software vendors face increasing difficulties in profiting just by selling software. Selling shrink-wrapped PC software has been enormously profitable: the cost of the physical package may be only $20 for a product that retails for $500. Support for this type of product is a cost center; and to be competitive, a vendor needs only to be just as good at providing support as the competition. Whoever holds the greatest market share can apply economies of scale to providing support and thereby gain a permanent advantage.

When Microsoft and Gartner coined the term “TCO,” profit margins on boxed software were huge; all you needed to make a fortune in the software business was to publish a useful, easy to use, and popular application that did not require significant resources to provide support.

Now, proprietary software vendors such as Microsoft want us to believe that cutting costs on software licenses will ultimately increase overall costs, and that the cost of the software itself is minuscule in terms of total IT-related costs. Software margins have to come down from the stratospheric levels of the 1980s, which means that vendors have to make up for it by selling more of the value-added services.

Increasing sales of a shrink-wrapped package means profitability will increase proportionately: one million copies of the package cost $20 million to manufacture, while ten million might cost only $150 million to manufacture.

Value-added services, on the other hand, don't scale well at all because they depend on an expanding pool of human talent rather than increased manufacturing capacity. Serving 100 customers who require high-priced services is far more manageable with 20 system engineers than serving 500 customers with the same 20 engineers. Coping with a five-fold increase in customers may require more than a five-fold increase in staff for a number of reasons. First, as the demand for skilled engineers rises, so does the cost of hiring those engineers (or training them in-house if the expertise is not widely available). Next, a group of 100 engineers means additional managers, administrators, and other support staff. Finally, positive efforts must be made to retain both customers and new staff as new capacity is added.

Obviously, this is a problem for some, but also an opportunity. As the role of human intelligence and skill dominates the total cost calculations of these systems, the service bureau model could be adapted for the new world of low-cost hardware and software. Rather than investing in machinery, the new service bureau startup would invest in people and skills. SMB customers could re-allocate their own internal staff to greater benefit while rationalizing their systems costs by turning over tasks to service bureaus.

Open source software makes this opportunity possible. Some open source projects are even designed around that opportunity, with the sponsoring company writing the software so they can use it themselves to offer a service. Inasmuch as open source licenses, by definition, can not restrict the use of the software, the only thing you do need to run such a business is some hardware and the skills.

If you've ever tried to install and implement open source business intelligence software, you'll know that support of some kind is almost mandatory. If an open source service bureau could be tasked with installations, as well as with customization and ongoing support/administration, many SMBs would undoubtedly show interest.

Whether or not the next wave of entrepreneurial activity includes open source service bureaus remains to be seen. The greatest obstacle to success for such a service provider would be in developing a reputation and track record for excellence before most SMBs would willingly hand over such an important part of their futures.

  • Pete LoshinPete Loshin

    Pete is Founder of Internet-Standard.com, an open source and open standard computing consultancy providing technology assessment, needs analysis and transition planning services for organizations seeking alternatives to commercial software. Pete has written 20 books, including “TCP/IP Clearly Explained” 4th Edition, Morgan Kaufmann, 2003) and “IPv6 : Theory, Protocol, and Practice,” 2nd Edition (Morgan Kaufmann, 2004).

    Pete can be reached at pete@loshin.com or at 781. 859.9175.

Recent articles by Pete Loshin

 

Comments

Want to post a comment? Login or become a member today!

Be the first to comment!