When I was a CIO, I always dreaded the annual budget season because I knew, somewhere during the process, the CEO, my boss, would ask, \u201cWhat are we getting for this constantly growing IT department.\u201d\n\nIt\u2019s a question that keeps most CIOs up at night when asked to defend IT investments, and it\u2019s one all CIOs should expect to answer, given that IT expenditures can range from 1% to more than 50% of a company\u2019s total revenue.\n\nFor most IT departments, this is a very difficult question to answer because the systems that we develop are not used by IT but are used by other departments to increase their sales, reduce their expenses, or be more competitive in the marketplace.\n\nAs such, an IT leader\u2019s usual response to this question is a general statement about how IT has implemented projects across the corporation that have achieved corporate strategic objectives. We seldom have any empirical data to back up our claims. So what\u2019s a CIO to do?\n\nIT as a business\n\nThere are two ways to address this issue. The first option is to transition from a non-charge out environment, where IT absorbs all development costs, to a charge out environment where all IT costs are assigned to the user departments based on their use of the resources. In this case, IT operates as a zero-cost department and there are no annual budget issues. All IT has to do is tell the user departments how much to budget for IT.\n\nBut there are great downsides to this approach that far outweigh its ease of use for IT. First, this process tends to place the automation agenda into the hands of individual departments or profit centers rather than looking at IT and digitalization as an overall company necessity. An example of this would be the development of artificial intelligence systems. The ramifications of this sort of system would affect all departments.\n\nSecond, with a charge out system, IT sends each department a monthly bill charging their P&L for a range of services, including development costs, IT infrastructure usage costs, and the dreaded overhead costs. This bill presents a huge challenge to IT to maintain its cordial relationships especially if it is higher than the budget estimates.\n\nPerhaps worse, shifting to charge out, or chargeback, approach treats IT like a business \u2014 a system that might sound good on the surface but means that the user department may begin to look to outside IT organizations to develop shadow IT systems that are sold as a cheaper alternative. These systems can only make internal system maintenance more complicated and drive a wedge through the company and its automation agenda.\n\nThe better way\n\nThe second and better way to approach the problem of IT value is to measure the effectiveness of the IT operation. Why should IT be the only department that is immune from corporate oversight? The advertising department is routinely measured on whether it is increasing corporate sales. HR is constantly being questioned on how its salary system compares to the industry. Manufacturing is always being challenged on its costs and if there are alternative methods and locations. Marketing must assure top management that its brand positioning is the best for the company.\n\nThe only way to measure IT is to enforce a requirement that all large scale new or modified system projects are analyzed, after completion, to verify that the objectives were met and the ROI was proven.\n\nIn my book The 9 1\/2 Secrets of a Great IT Organization the 1\/2 secret is the post-implementation audit. I called it a half secret because few companies do it. It should be treated as a full secret, however, because it will assure a much more effective and successful department. But it is not generally done for a number of reasons.\n\nFirst, conducting a post-implementation audit requires a significant amount of analysis that is very detailed and can span several years. Just gathering the data can be time consuming especially given that many of the project personnel may have changed jobs or even companies since the project was completed.\n\nNext, it cannot be done until at least a year after the system has gone live since no system is fully functional on day one. Sometimes it is hard to convince both IT and the user department that it is worth the time to analyze a completed system because there are more important projects to complete.\n\nMoreover, the user department is often not interested in proving ROI for several reasons. Perhaps they inflated the initial ROI to get the attention of the IT steering committee. A close analysis may discover this practice. Additionally, the ROI may have contained significant headcount reductions that were used to generate a better return. The department may desire to forget these moves once the project is completed.\n\nOf course, it\u2019s not always about the user department. IT may also not want to see the audit done because it may have underestimated the cost or completion date on the original estimate.\n\nThe recommended way to complete this audit is to remove the responsibility from the user department and from IT. An independent organization, preferably under the auspices of the financial arm of the company, should conduct the post-implementation audit. This group should have been involved initially in developing the ROI for the project and are in the best place to assure objectivity in the result.\n\nIf done this way, the user department will be held to its ROI commitment, IT will be held to its performance objectives, and the CIO will be able to answer the question posed by the CEO about IT investments by saying, for example, \u201cWe implemented 17 projects this year which increased sales by 35% and reduced expenses by 14%.\u201d \n\nWouldn\u2019t that be a great conversation to have, not only with the CEO but with the entire company.