How many IT services vendors do you rely on?\n\nSplitting responsibility for the IT organization into multiple outsourcing vendors, overseen (or overlooked in some unfortunate cases) by a small IT management team, has become a popular practice. Hardly \u201cbest practice\u201d \u2014 a meaningless but popular justification for doing things a certain way \u2014 but popular nonetheless.\n\nIf you\u2019re on top of an IT organization that\u2019s structured like this, or if you\u2019re thinking about joining the club as a bold way to turbocharge IT organizational performance, don\u2019t be hasty. It\u2019s an alternative that can work, but it can easily backfire.\n\nHere\u2019s a sampling of what can go wrong and what you can do to prevent it, or at a minimum mitigate the risks.\n\nService level addiction\n\nYes, yes, yes, if you can\u2019t measure you can\u2019t manage. That\u2019s doubly so when assessing a vendor\u2019s performance.\n\nAnd yet \u2026\n\nWhen you\u2019re leading an internal team there are intangibles you insist on \u2014 for example, you might ask for innovative thinking and a willingness to go a few extra miles to turn innovative thinking into innovative reality.\n\nSLA-driven vendors can become complacent, and in the absence of a well-defined and tracked \u201cinnovation SLA\u201d neither they nor their management will have any incentive to suggest anything new.\n\nIt\u2019s tricky. On the one hand, the vendor you select should, in theory, have more expertise in their area of responsibility than an internal team might have, and so it should be able to offer a steady stream of improvement possibilities.\n\nBut on the other hand, many \u2014 perhaps most \u2014 improvement opportunities would reduce the vendor\u2019s billings. It\u2019s hardly reasonable to ask a vendor to take the initiative in reducing their own revenue, unless you can at least offer them increased margins or some other incentive.\n\nLeadership voids\n\nLeadership is about people and motivation. Management is about getting work out the door. In a very real sense, outsourcing is about shifting the CIO\u2019s attention from leading to managing.\n\nThat\u2019s a mistake. An outsourcer\u2019s employees should certainly receive day-to-day leadership from the vendor\u2019s account manager. But that\u2019s in addition to the leadership they should get from their client\u2019s CIO.\n\nThey are, after all, people, too.\n\nInter-vendor politics\n\nJust as there\u2019s no such thing as a perfect org chart \u2014 one that clearly and unambiguously defines each manager\u2019s responsibilities in ways that avoid overlap \u2014 so there\u2019s no such thing as a division of outsourcer responsibilities that avoids the potential for game playing.\n\nPotential? The game playing is often overt, although concealed from view: Many vendors have regular internal strategy sessions in which they exchange ideas about their \u201cendgame\u201d \u2014 how they plan to eliminate a competing vendor from the IT organizational gameboard.\n\nThe most common examples of inter-vendor politics occur when one vendor needs data from another vendor in order to move a project forward, and the vendor that receives the request, instead of providing the asked-for data, emits an unending stream of excuses, or complains that providing it isn\u2019t in the contract, and as a custom service it will cost a significant and unbudgeted sum.\n\nOften, the stonewalled vendor will be hesitant to complain, on the grounds that this might look a lot like they\u2019re playing inter-vendor politics themselves.\n\nAnd there\u2019s another layer of complexity to deal with when trying to prevent, or at least manage, inter-vendor politics: In addition to one vendor trying to torpedo a rival by failing to fulfill data requests, there are times a vendor will request data from a rival in order to uncover something damaging about that vendor\u2019s performance.\n\nThe annoying solution to inter-vendor politics is to insert yourself into any and all inter-vendor interactions. Because Vendor A refusing to provide information to Vendor B is one thing. Refusing to provide it to you? It\u2019s your data.\n\nExit mitigation migraines\n\nHere in the USA our entire system of economics is built on a single empowerment, namely, that a customer can threaten to take their business elsewhere. When it comes to IT services vendors, however, this can be a hollow threat.\n\nTo be effective, an IT services vendor\u2019s employees have to engage in the osmotic process of learning their customer \u2014 its key staff and their quirks and temperament just as much as the application portfolio and integration architecture.\n\nOnce a services vendor is firmly entrenched, transferring that incumbent\u2019s tribal knowledge to a replacement vendor would be, to put it gently, a non-trivial task, and that\u2019s just the quantitative view. The political dimension exacerbates the problem: Why would you expect a vendor you\u2019re kicking to the curb make it any easier than necessary for their replacement to succeed?\n\nFair\u2019s fair, though. The same was true for the IT staff the CIO kicked to the curb during the outsource.\n\nAnd beyond the political angle, in many cases the services vendor will have installed its own toolkits to help in fulfilling its responsibilities. It will take that toolkit with it should you decide to make a change.\n\nSo on top of all the rest of your exit mitigation planning, make sure your contract includes an obligation to leave these toolkits in place for a long enough transition that the replacement vendor has the time it will need to install its own proprietary toolkit.\n\nIn any event, prudent CIOs have two possible exit strategies, and they\u2019re complementary, not dichotomous.\n\nThe first: Keep a small cadre of IT professionals on staff to work alongside each vendor. That way, if circumstances dictate, they can smooth the transition to the next vendor.\n\nAnd the second? Keep a close enough eye on your vendors that you can circumvent the need to replace any of them in the first place.\n\nIt\u2019s a quandary. Managing outsourcing vendors is arguably more complicated than in-house management and staff, because CIOs who have outsourced IT services have fewer management tools at their disposal than those who need to lead employees.\n\nAfter all, when you have to deal with poorly performing employees you can call on HR.\n\nBut when you\u2019re dealing with a poorly performing vendor, in contrast, who do you have to call on?\n\nThat would be your company\u2019s general counsel, who would be no happier about being called in than you\u2019d be in calling them.