Apples and Oranges: Comparing IT Maintenance Costs to New Initiatives

By Jan Vromant

(Back to article)

Most IT departments spend as much as 80% of their budgets on routine maintenance and day-to-day operations, while only 20% is spent on new technology or business-process enhancements, according to surveys by Gartner and other research firms.

The conventional wisdom is, with the total IT budget being constant, it is better to spend more on new initiatives and minimize spending on legacy systems. This way of thinking has some important downsides, mainly because the inherent nature of maintenance costs and “truly” new initiatives is as different as apples and oranges.

Here are some examples of conventional thinking I’ve encountered:

  • In a major Midwest financial services firm, the key driver in planning and budget discussions is a ratio of 70% for new initiatives and 30% for maintenance costs. There is considerable pressure on the leadership team to achieve this and it forms part of their performance review.

  • Another example came up in a recent conversation with a partner at a major consulting firm. He had been working with a Fortune 500 company on the West coast, and the CIO asked him for suggestions on how to shift the balance from their present 75/25 or so to a more “forward-thinking” ratio.

  • Further digging into the topic brought an example from the CIO of HP: "... Randy Mott's report to the Hewlett-Packard board of directors [...]: Shift 80% of staff resources to new projects and focus only 20% on maintenance--all by 2009. Mott's goal represents a reversal of the standard 80/20 rule and a considerable improvement over the HP IT organization's current 54/46 ratio of new/legacy IT work." ("CIOs Uncensored: HP's Mott Goes For Broke, Seeks 80/20 Reversal" - Stephanie Stahl, InformationWeek, 27 Nov 2006)

    As a post-facto informational ratio, the ratio between maintenance and new initiatives can be useful, because it highlights where the net maintenance costs are heading. It might also be a driver in a decision to outsource some legacy systems. However, this ratio should be used with caution and not as the main driver for budget discussions. The distinction between maintenance costs and new initiatives should be guided by the business needs, within an IT services lifecycle framework.

    Poor Tool

    The concept of a ratio between maintenance and new initiatives is easily understood and hence attractive, but it is a poor tool to manage your IT portfolio.

    The main reason is that new initiatives and investment proposals should be judged on their own merit and not be subjected to a push caused by a blind chase for a ratio. Common sense dictates that each investment proposal should be looked at individually and in conjunction with the whole IT strategy and portfolio of business functionality and IT architecture.

    Each major proposal, each major change, each investment needs to be able to stand on its own in terms of rationale and return on investment. The desire to shift the legacy- /new-initiatives ratio should not be a factor in the approval of an investment. Using the ratio as a tool to guide your budget can lead to a major misalignment with the business.

    There are some additional arguments to be made against using this ratio:

    The definition of the split is akin to nailing pudding to the wall and can hence be too political. If you are planning an upgrade to your SAP environment, do you consider the cost a part of the maintenance or is it a new initiative?

    If you are embarking on a major new project setting up ITIL processes, is that part of putting your house in order and a legacy cost, or is this something “new”? A friend in the aforementioned Midwest financial services company puts it as follows: "... the issue we face … is what qualifies as a new initiative and what qualifies as maintenance. The line is fuzzy and gray so it makes it hard to truly achieve the ratio."

  • In addition, the reason of shifting projects from the category of maintenance to new initiative can be inspired by a desire to move budget money around—a new initiative might be considered a capital investment and does not immediately show up as an expense. An easily manipulated metric is a political metric.

    The metric mainly depends on the IT architecture and build-up history of IT. New initiatives are seldom totally “forward-thinking” and really innovative. Most of the time, your new initiatives will not be implemented in a vacuum, but will be a layer on top of an existing infrastructure.

    A company with a massive ERP, e.g., SAP, will have a fundamentally different and higher ratio than a company with multiple applications. The company with the ERP will spend time optimizing this ERP, which is considered a maintenance cost. The company with the multiple applications will invest in new applications to fill business functionality gaps, which are considered new initiatives.

    This makes it impossible to compare the ratio in one firm with the ratio in another firm and reduces the usefulness of the ratio as a benchmarking tool.

    Business doesn't care about the ratio, as long as they know that, a) the total IT budget is under control; and b) they get the services they need. IT is becoming more and more a commodity, as Harvard's Nicholas Carr pointed out in his article "IT Doesn't Matter". The ultimate question should be about the ratio effect on the business functionality and services offered, as they are the reason of existence of IT. Changing the ratio does not link to a direct or obvious benefit for the business.

    The net amount of the costs for the legacy systems should decrease year-over-year, regardless of new initiatives, for two reasons. First, Moore's Law is ubiquitous in IT: the cost of networks, servers, hardware should go down over time. We all know the painful experience of buying an electronic item and then seeing it advertised three months later for 25% less.

    Second, the IT department should get better at managing its environment and using its known error database and gradually reducing the total number of recurring incidents. This feeds into the total cost of ownership (TCO) management, which should go down because of the gained experience.

    The main source of IT's headaches are changes. New investments and major changes should only be pursued when the environment is mature and stable. New initiatives for the sake of being more forward-thinking can be disastrous. IT departments are already buckling under the yoke of more and faster changes every year. New initiatives could take energy and time away from the essential goals of providing a stable IT service to the business and enabling maximum business productivity.


    The important aspect is not to blindly follow a hybrid and pleasantly easy-to-understand ratio of IT investment versus IT maintenance expenses, but to rationally dissect the purpose and usefulness of this ratio and to compare apples with apples, and oranges with oranges lest we wind up with some hard-to-digest fruit salad.

    Jan Vromant is a process architecture consultant, focusing on outsourcing and ITIL best-practices and is a lead in the Outsourcing Advisory Services of Deloitte Consulting, LLP. Before joining Deloitte, Jan worked at Royal Dutch/Shell, BMC Software, PricewaterhouseCoopers Consulting, and Hewlett-Packard.