Effective IT organizations rely on these foundations: well-defined, -designed, and -implemented integration with the enterprise; sophisticated process and practice oversight; and a robust, up-to-date technical architecture.\n\nBut without the right people, pointed in the right direction, your IT efforts will be futile, no matter how good these look in the documentation. A CIO\u2019s success, that is, depends more on human performance than all other factors combined.\n\nNor is the question whether individual humans in your IT department are performing. It\u2019s about the factors within the CIO\u2019s control or influence that foster strong employee performance if handled well, or inhibit it if handled poorly.\n\nCIOs have three tools at their disposal to encourage employees to perform at their best: compensation, organizational structure, and leadership. One of them even works reliably.\n\nCompensation\n\nLet\u2019s get this out of the way: Paying people more won\u2019t result in better performance.\n\nImagine you\u2019re interviewing a software quality assurance analyst. In the interview you ask, \u201cWhat would you need from me to minimize the number of defects at each stage of our software development lifecycle?\u201d\n\nThe applicant\u2019s response: \u201c125,000.\u201d\n\n\u201cI beg your pardon?\u201d you say, nonplussed.\n\n\u201cIf you pay me $125,000, I\u2019ll minimize the number of defects,\u201d the applicant explains. \u201cFor $100,000, I\u2019ll keep the number small. $75,000 buys you software that won\u2019t crash your production environment.\u201d\n\nYou wouldn\u2019t hire an applicant who pegs their promised performance to the amount you pay them. Employee effectiveness can\u2019t, that is, be obtained through bribery. And yet, many managers think they can effectively use compensation as a performance incentive. They\u2019re wrong. Paying too little can be a powerful disincentive, but that\u2019s a different matter.\n\nCompensation is a terrible incentive, due to the phenomenon ethologists call \u201chabituation.\u201d Habituation is getting used to something so its impact declines with repetition.\n\nGive someone a raise or bonus today, and from here on in, getting another one is expected. Additional compensation stops being an incentive and instead becomes an entitlement, which, if it fails to show up next time, becomes a disincentive.\n\nYou avoid the incentive habituation trap by recognizing compensation\u2019s proper role: It puts the company\u2019s money where its metaphorical mouth is.\n\nCompensation, that is, is the company\u2019s loudest voice for making what it values clear.\n\nIn the human performance context, compensation falls into three buckets: base compensation, spot bonuses, and the annual bonus.\n\nBase compensation: Set this so no employee has a financial incentive to leave for greener pastures. Annual raises, in consequence, communicate how much more you think each employee is worth in the labor marketplace now than they were a year ago.\n\nDon\u2019t change base compensation based on an employee\u2019s performance over the past year. That would make a raise an annuity, which you\u2019d pay out in the future whether or not the employee continues to perform.\n\nAnnual bonus: This is how you provide compensation for performance \u2014 for contributing exceptional value to the organization over the past year.\n\nThe annual bonus is a very loud voice. It clearly explains what constitutes exceptional performance, and the extent to which an employee exhibited it over the past year. It\u2019s loud because you can, in round numbers, break even financially if an employee\u2019s annual bonus is three times as big as their annual raise would be if you used that to reward performance. This works because unlike an increase in base compensation, the annual bonus isn\u2019t an annuity.\n\nEmployees have to earn it all over again next year.\n\nSpot bonus: When an employee does something that\u2019s clearly above and beyond the norm, a spot bonus is in order. But don\u2019t give them the money as a reward. Also don\u2019t give them money as an incentive to everyone else to go above and beyond. That will trigger the habituation trap.\n\nNo, from the loudest-voice perspective spot bonuses handled properly deliver a simple and sincere message: \u201cThank you!\u201d\n\nThat\u2019s a good message to deliver.\n\nAt the end of it all, what\u2019s most important about how you handle compensation is that employees perceive it to be fair. Fairness won\u2019t lead to their working harder and with more dedication, but perceived unfairness will definitely dissuade them.\n\nStructure\n\n\u201cIt must be April,\u201d a friend observed. \u201cIT is reorganizing again.\u201d\n\nReorganizations rarely improve an organization\u2019s performance. Quite the opposite \u2014 they impair it. They\u2019re popular because they buy time when the executive leadership team calls IT\u2019s performance into question while avoiding the risks associated with deep and fundamental change.\n\nA reorganization is an exemplar of Titanic Deck Chair Rearrangement Syndrome. It keeps the same work groups in place and makes them responsible for the same work, altering only which work groups report to which managers.\n\nSo far as how work gets done, nothing changes. Meanwhile, employees and managers who knew how to work together no longer work together.\n\nThe most frequently claimed benefit from reorganizing is that it can reduce the height of barriers to getting decisions made and work done by reducing the \u201corganizational distance\u201d between groups that have to work together. But the benefit is temporary, as reorgs raise as many barriers as they lower.\n\nOften, the purpose of an IT reorg is to \u201cflatten the organization\u201d \u2014 to reduce the number of layers separating the CIO from the people who do the actual work. Fewer management layers means less of what staff know that the CIO needs to know gets filtered out.\n\nFlattening can and does work. But it\u2019s not an unmixed blessing. For every layer that\u2019s eliminated, the manager at the next layer up has that many more direct reports, and therefore has that much less one-on-one time to spend with any of them.\n\nWhen the goal is to improve a CIO\u2019s ability to know what\u2019s really going on in the IT organization there are other, better tools at their disposal \u2014 tools like \u201cskip lunches,\u201d anonymous employee surveys, open door policies, along with metrics and performance reporting \u2014 that are far less disruptive, and usually provide better and more unfiltered information.\n\nChanging the organizational structure rarely fixes anything that\u2019s broken, but it has a lot of potential for breaking what\u2019s currently fixed.\n\nLeadership\n\nThe word \u201cleadership\u201d has a mystique to it. It\u2019s shrouded in the biographies of great historical figures, compounded by our experience of charismatic presences in our lives, and further obscured by the occasional snappy management guru quote.\n\nThen there\u2019s the discouraging assertion that \u201cyou can\u2019t teach people to be great leaders.\u201d This truism is an excuse that obscures the more useful perspective: While being a great leader is beyond the reach of most people, becoming a better leader depends on entirely learnable skills.\n\nBut this starts with understanding what leadership is \u2014 its definition. Leadership is the art of getting others to follow. If they\u2019re doing that then you\u2019re leading; otherwise, you aren\u2019t.\n\nElsewhere I\u2019ve enumerated the \u201ceight tasks of leadership\u201d (Leading IT: <Still> the Toughest Job in the World) \u2014 specific techniques anyone can learn that can make them better leaders. From the perspective of improving the performance of the IT organization you lead, what\u2019s most important is recognizing the single most important fact of effective leadership: that effective leaders don\u2019t get anything done. They build organizations that get things done.\n\nThe best organizations, that is, follow from in front.\n\nThe best leaders, in their turn, make sure everyone in the organization they lead knows where \u201cin front\u201d is, agrees that moving in that direction matters, understands the role (or roles) they have to play in moving the organization forward, and commits to that role.\n\nThe best leaders also encourage everyone in their organization to think of themselves as leaders regardless of their title, providing leadership to those around them so they also know where \u201cin front\u201d is, understanding and committing to their role in getting there.\n\nAnd in conclusion\n\nEverything you have to do to improve human performance takes time \u2014 time you probably don\u2019t have. Fix this first.\n\nAnalyze your calendar. Tally how many hours you spend in meetings and undertaking tasks someone else put on it, compared to how much of your time is self-directed. If you don\u2019t spend at least four hours a week in self-directed activities you have little chance of improving your organization\u2019s human performance dimension.\n\nSo the first step in improving your organization is getting control of your calendar. This will be hard, because most of us, most of the time, spend more time working on what\u2019s urgent than on what\u2019s most important, relying on others to let us know what\u2019s urgent.\n\nOn any given day your human performance responsibilities probably won\u2019t be what\u2019s most urgent.\n\nBut to lead a highly effective IT organization, nothing is more important.