Arthur Andersen studies sponsored by Microsoft notwithstanding, the history of TCO and ROI analyses indicates that they are still accepted by planners as guides, however imprecise. This is a continued business climate where information managers are under the gun in justifying their computing constructs. TCO analyses help support their decisions. ROI analyses, as their names suggest, correlate returns to investment.

Total Cost of Ownership (TCO) is an assessment that helps an enterprise CFO to understand how much financial resource will be eaten up if a new piece of technology is brought into the business. A TCO study shows the direct and indirect dollar costs of owning and using any IT component throughout its lifecycle. TCO data attempts to cover all the phases such as acquiring, installing, and managing a company’s computers, networks, and applications.

Who invented this “TCO” acronym that is now a vendor, analyst, and user term, anyway? The credit goes to the Gartner Group, which first introduced its total cost of ownership idea over five years ago. In 1996, Gartner won a lot of business attention to the hard and less apparent costs of IT when it said that it cost at least $13,000 for a corporation just to maintain a workstation.

The following year, Gartner went on record as making IT managers think hard if they were really getting their money’s worth with Microsoft, when it estimated the TCO of a Windows 95 PC as $9,784. Then in 1999 Gartner went on record presenting TCO figures to show corporate managers that the TCO for Windows 2000 was imprudently high. That’s when TCO entered a most interesting stage, with Microsoft issuing counter-reports, and the Giga Group issuing its economic analysis findings (under its own TEI moniker), showing Windows 2000 benefits nevertheless outweigh the costs of migrating to Windows 2000.

Everyone was talking TCO. CIOs sought TCO justifications to bring before their CFO and CEO arbiters; vendors sought TCO studies to prove that their products and platforms made better sense than that of competitors; analysts beyond Gartner were promoting their independent studies of TCO with clients. The Radicati Group, for example, leveraged messaging technology expertise in preparing detailed TCO studies showing cost comparisons between using Lotus Notes and Microsoft Exchange.

What are the obvious and soft costs, TCO studies ask, that surface beyond the mere price of the hardware and software? What about support, what about administration, what about ease of learning, what about downtime, what about upgrades, what about change management?

More importantly, are the answers really that useful? One engineer complained in a Web feedback session that TCO authors “futz” with numbers and paint scenarios that may be irrelevant for some decision makers. To counter that TCO-is-too-full-of-guesswork argument, analysts now hasten to show the business and vendor world that they have extra-strength assessment approaches that go beyond vanilla TCO.  Advisory groups and vendors of all sizes and areas of concentration have found a way to leverage the fact that averaged-out TCO white papers may suffer from imprecision that won’t serve the executives well as their mainstay guide for technology buying. These groups say they have something better.

One promoted gift to the business gods is the Giga Information Group’s TEI (Total Economic Impact) approach, which Giga positions as good both for enterprise buyers and for technology vendors that are out to build a case for their solutions. Giga’s Stacey Quandt, Industry Analyst, Linux and Open Source, says: “Total Economic Impact is a holistic ROI approach that measures how a solution or initiative not only impacts IT but also the company's business units.” As a result, the TEI methodology includes factors for Flexibility, Benefits, Costs and Risk. That way, Quandt maintains that study recipients get a better picture of the true business value of a technology initiative or solution. “TEI not only measures how a solution impacts IT but also company business units. It expands traditional cost analysis to include calculations on excess capabilities and ‘risk-adjusted’ returns.”

Microsoft, dwelling on its beloved business at the rate of Internet speed theme, says that IT managers need a quick approach to understanding the value of IT investment, can’t afford to wait around for months before a study appears, and can benefit from something called the REJ (Rapid Economic Justification) framework, which is more business-centric.

The Standish Group International (West Yarmouth, MA) used to be content with TCO nomenclature until they came up with something extra: Its “TCO in the Trenches” report of 1999 that outlined high-end server costs has been updated this year and also renamed to Dollars to CENTS: TCO in the Trenches 2002. Standish Group’s CENTS stands for Comparative Economic Normalization Technology Study. CENTS cases are built using data input by users. No vendor data is used. The CENTS cases drive another mouthful, the group’s VirtualADVISOR Cost Assessment Model database.

What does all this have to do with Linux?

IT managers trying to answer their bosses’ requests to look into Linux as a reasonable next-step are looking for assessment answers. They won’t find it merely in conversations, which will turn up confusing misconceptions. New enterprise-level potential adopters of Linux are confronted with pro-proprietary, anti-Linux, propaganda full of dire warnings about poor security, the risk of being an administrator stuck with no proper support, and the risk of bearing the costs of a future full of system complexities that will need more staff.

Microsoft claims that Windows NT has a better TCO outlook than Unix. Then by extension, the argument follows that this advantage extends over Linux. Microsoft’s 1999 white paper focuses on comparisons with NT 4.0 Server and presents Total Cost of Ownership (TCO) research showing Unix-based systems having a 37% higher TCO than NT. The assessment compared Windows NT Server running on Compaq hardware against Solaris running on SPARC hardware. The page presenting that study has since gone away. Nonetheless, a rebuttal to the Microsoft argument is still on the web.

Later on, a Microsoft “Linux Myths” page, said that the cost of the operating system is only a small percentage of the overall total cost of ownership (TCO). In general, said the report, Windows NT has proven to have a lower cost of ownership than Unix and that previous studies have shown that Windows NT has 37% lower TCO than Unix. “There is no reason to believe that Linux is significantly different than other versions of Unix when it comes to TCO. Linux is a Unix-like operating system and is therefore complex to configure and manage. Existing Unix users may find the transition to Linux easier but administrators for existing Windows-based or Novell environments will find it more difficult to handle the complexity of Linux. This retraining will add significant costs to Linux deployments. Linux is a higher risk option than Windows NT.”

The bombshell argument against these claims comes in the form of an International Data Corporation (IDC) study sponsored by Red Hat that quantified the TCO benefits of running Linux compared to Unix. IDC analysts Al Gillen, Dan Kuznetsky, and Scott McLarnon concluded that the costs associated with Linux were dramatically lower for the hardware, software, and more importantly lower for staffing—which justifiably raised a lot of eyebrows because it went against the grain of popular thinking, and contradicted gloomy warnings from Linux detractors.

“With staffing typically the largest component of overall IT solution costs, this finding has important implications for IT planning,” says the study.

Another important study looking at Linux vs. Microsoft NT in the back office was done for the U.S. Department of Defense conducted by MITRE Corp., which says Linux was easier to manage, had more robust security features, and supported remote monitoring and management more effectively. MITRE found that each of these features resulted in measurable cost savings and risk reduction.

Quandt of Giga Group, meanwhile, argues in favor of using Giga Group’s TEI approach for considerations in deploying enterprise Linux: “Some of the most harmful mistakes we have witnessed,” claims Quandt, “include targeting costs savings rather than Total Economic Impact of a Linux infrastructure decision.”

Examples of what she sees as shortcomings: “Some TCO studies present an abstraction of real-world lifecycle costs by allocating all server hardware and software to the first-year costs.” She says that another shortcoming is that the TCO analysis does not always detail specific ISV applications. “The findings of these TCO studies usually focus on the benefits of migrating from Unix/RISC to Linux/Intel but fail to mention which applications are involved, nor do they reflect the implications of a build vs. buy decision.”

Don’t expect this steady diet of TCO proofs/counterproofs to end any time soon.

http://www.fortuitous.com/nut/cost.html
http://sunsite.dk/linux-newbie/lnag_benefits.html#TCO
 

Ditto over debate as to whether total cost of ownership proves anything. The Gartner Group makes the inarguable observation that all CEOs and CIOs are looking for IT projects that reduce expenses and increase efficiencies. Yet “The total cost is not what matters the most,” asserts Linux Newbie Administrator’s Guide:  “What value did I receive for my money? You would have to calculate the total value of ownership (TVO?), then subtract from it the total cost of ownership (TCO) to obtain the ‘net benefit of the ownership.’”