TCO is an extremely useful accounting system to help IT teams drive better purchasing decisions by considering the total economic impact of each decision. The underlying driver is that the solution with the lowest purchase price could, in the long run, cost more. TCO analysis enables teams to compare various purchases and options to select the solution that will cost the organization the least over its lifecycle.
In order to appreciate the total cost of an asset, TCO analysis considers costs such as change costs and on-going operations management, administration and support labor. These can often be staggering, representing some 70% of the overall direct cost of owning an asset; labor and equivalents often consume more than 56% of the total cost.
Surprisingly, only 15% of average total IT costs are for servers and client computers, and only 18% for purchased software. Yet, how many purchase decisions within your IT organization are made solely on the basis of lower hardware and software costs? If the majority of them are, you are not alone as most organizations are focused on purchase price rather than the on-going annual expenses to manage and support the asset.
As well, computer assets can sometimes have hidden costs such as the impact of security issues or poor service levels, causing downtime and business losses. A
The TCO metric was developed specifically to measure the true cost of computing and highlight these issues. TCO's roots go back to the late 1980's, when distributed systems were being implemented en-mass because they were perceived to be much less expensive than prior data center computer systems. In reality, there were dramatic up-front cost advantages, but these were much less than expected because of overall on-going and hidden costs. These studies, particularly those from Bill Kirwin, the father of TCO at Gartner, determined that the purchase price of the PC hardware and software was only 15% of the total cost of owning the asset. Management, support and hidden support and downtime costs accounted for 85% of the total cost over the useful life of the asset. At the time, a PC that cost $2,000 to $3,000 might actually cost the organization over $8,000 per year or more to keep in service. Before TCO, many IT executives and even solution providers were unaware of the true cost of computing. TCO made every one poignantly aware of the issues.
TCO proved to be an excellent tool that analysts could use to highlight that manageability and availability of systems was important. Reducing these costs should be a focus of solution providers and IT professionals alike. With more manageable systems, costly systems administration and support costs could be reduced, as well as reducing costly shadow support. With greater reliability, support costs could be further reduced and downtime impacts could be eliminated. TCO trained teams to look beyond purchase costs alone and to focus more on manageability and hidden costs.
However, TCO came under criticism because it placed little on the value of the asset towards meeting business goals. As many would say, to lower TCO we could all resort to pen and pad, which has a TCO of $1.50, compared to an estimated $7,000 per year for the typical Windows computer system. Rightly so -- by focusing on costs alone, the business benefits of proposed purchases could be overlooked.
So when is TCO most useful? First, if the team does not have a clear understanding of what solutions to consider, a quick TCO analysis of current costs can help determine where they may be too high, or where service level issues are costing the business. This is especially useful if the team can compare the current company TCO to those of peers and competitors. With specific knowledge of high costs or poor service levels, the team can prioritize various practice, capability and maturity improvements.
When the team reaches consensus that a particular solution is needed, such as a new business application, and there are several similar solutions that might work at delivering the same, or similar benefits, TCO is helpful to compare the cost of these solutions head-to-head to determine which solution will be most cost-effective over its useful lifecycle.
TCO is less effective at comparing solutions that are dissimilar. Often, vendor marketing campaigns use TCO in a less than credible manner to highlight competitive advantages without comparing solutions that deliver similar business benefits or capabilities. For example, our studies of server TCO reveal that if workloads are not precisely aligned for each solution being compared -- or the optimal number of servers are configured to support the performance and service levels needed -- results can vary by 20-40% or more.
TCO analysis also requires that a proper and complete cost accounting be performed, that captures all costs for each solution. More often than not, various TCO studies ignore critical cost factors and do not use a complete chart of accounts, leaving out critical costs such as tallying all change costs, administration and support labor, security risks, downtime and shadow support.
Projects for which TCO can help make the best decisions include:
- Computing platform and operating system (Red Hat Linux vs. Microsoft Windows Server 2003, or blades vs. rack mount)
- Database and application platforms (Oracle vs. Microsoft SQL) Build vs. buy
- Manage your own vs. managed services vs. hosted solution
Tom Pisello is the CEO of Orlando-based Alinean, the ROI consultancy helping CIOs, consultants and vendors assess and articulate the business value of IT investments. He can be reached at firstname.lastname@example.org.
This was first published in September 2005