Gartner made a big splash late last year with its prediction that by 2017 the CMO will spend more on IT than the CIO. Likewise, IDC predicts that what it terms "line of business executives" will control 40 percent of IT spending by 2016.This generated a lot of discussion, as in this example from Gigaom. Most of the discussion focused on predictable topics: The relationship between CIOs and CMO (not very good), what type of apps CMOs focus on (very interesting, and something I'll address below) and how vendors need to change the way they go to market when selling to a CMO (everyone thinks it will be different, but nobody's sure exactly how).(I think you can generalize from CMO to the line of business executive, or LOB, and take as a given that they have similar goals and objectives. In the discussion below, when you read CMO, think CMO\/LOB.)It's interesting that nobody seemed to comment on the type of IT spending that this shift portends. It was, in effect, as if everyone assumed the pie, though cut differently, but it would be the same kind of pie.Analysis: Does the Rise of the CMO Threaten CIOs?This is dead wrong. Everyone needs to rethink their assumptions about IT budgets when considering how they're spent by CIOs and CMOs. It is crucial to understand what this shift means for the relationship between IT CapEx and OpEx, and it's especially important to understand what the difference this means regarding the metrics the two roles (CIO and CMO) are measured by: Total cost of ownership (TCO) and return on investment (ROI).What should we think about this prediction that CMOs will come to control more than half of IT budget? There are several obvious implications\u2014and, I should note, they are true whether CMOs control 25 percent, 50 percent or even 75 percent of the IT budget. It's vital to understand how the dynamics of IT spending are going to change in a world in which CMOs control a significant portion of the IT budget.Less Money for the CIO Means Less Control, CloutFirst, and most obviously, if you're a CIO who controls less of the overall budget, you have less clout within and outside the company. To vendors, you are more or less interesting based on how much you might spend. If you have less money, you're less interesting.One can predict that CIOs, in particular, and IT organizations, in general, will move to an influencer role\u2014meaning they'll affect the evaluation but won't make the decision, or hold the purse strings. A corollary is that CIOs and IT organization personnel will undergo an uncomfortable learning process as they shift from frontline decision-maker to part of the supporting cast.The shift of attention to CMOs will cause change for vendors as well. Vendors can expect more emphasis on business outcomes and benefits, with less focus on standards, functionality and feature checklists.Less Visibility Means Spending UnpredictabilityFar more important than the "who's up and who's down" of buying and selling politics will be what affect this new buyer has on IT spending patterns. In other words, the who affects individual deals\u2014"I'm vendor X\u2026who is the decision maker for this purchase?"\u2014while the what affects the makeup of the entire budget. It's here that the new decision-maker and next-generation application profiles really change the nature of IT spending.Analysis: 6 Ways CIOs Can Make Peace with CMOsThis new CMO cloud world will cause IT spending within a company to be far less predictable for the following reasons:1. In the old, IT-controlled world, yearly budgets were more or less set in stone by Jan. 1. Beyond the obvious fact that headcount costs were more or less fixed and could be forecasted by that date, most other spending was fixed as well. Because nonmaintenance, nonheadcount spend was primarily capital expenditure for long-lived assets, it had to go through long planning and budgeting exercises before the year it would be spent, often two or more years in advance. It's common for IT organizations to be researching something in one year, putting it into the budget planning the next year and spending the following year.In the new, CMO-controlled world, much of the spending will be focused on public cloud computing, where no capital investment is required, but where all costs will fall into an OpEx category. The fact that this spending does not require CapEx means that less visibility into budgets will be possible on Jan. 1 of a given year. 2. This unpredictability will be exacerbated by the fact that these CMO-oriented applications, by their very nature, have much higher variability of load and user base. If your application is tied to a movie's release, and the star of the movie experiences a personal scandal that drives enormous interest in the movie, your app may experience 10 times the traffic you expected\u2014and 10 times the cost for your cloud computing.3. A third factor increasing the unpredictability of IT spending lies at the intersection of the decision-maker profile and cloud computing itself. CMOs frequently are presented with new business opportunities that don't fall neatly into "beginning of the year" launch timeframes. If it's a good opportunity, you want to pursue it right away, not wait for the next budget cycle.ISACA: Lock-in, Migration Costs Can Put a Damper on Cloud ProjectsIn the past, when you had to budget capital to pursue an opportunity, the process meant the increased spending was quite predictable, as it had to be planned for and slotted into the yearly budget. Today, when cloud computing makes resources available in minutes, there's little to prevent you from starting right away. Even if there's a budget forecast at the beginning of the year, you can expect significant variation as CMOs choose to pursue promising new business opportunities.In short, the new world moves from predictable CapEx to unpredictable OpEx. You can expect many more midyear budget recalculations as cloud-based application costs move all over the map.ROI, Not TCO, Drives OpEx SpendingTCO is the mantra of IT organizations. IT is a cost center, and it's a given that the focus will be on keeping costs down. The less you spend, the more that's available for the overall company's profit.Consequently, IT organizations negotiate to keep costs to a minimum and are loathe to consider more expensive alternatives. After all, from an IT perspective, spending more on one product means having less to spend on another. This is particularly true in a capital-rationing environment.If IT is allocated only a certain amount of capital to spread across all capital needs, then it's going to try and trim spending in every area to stretch the capital as far as possible. Generally, the most effective way to sell to a CIO is to offer a way of doing something cheaper than it's being done now, because that will reduce TCO.CMOs, by contrast, focus on ROI. CMO spend is bound to be focused on measurable business outcomes, and the applications they use (or have built on their behalf) are directed toward achieving those outcomes.For a CMO, the question is, "If I spend an IT dollar, how much do I make?" Even more important, "If I spend a dollar, what is my margin on that dollar. Is it the same as the previous dollar I spent, less or more?" The business case for a CMO is simple: If I can make as much or more margin by spending an additional dollar, I'm going to spend it. Moreover, if achieving that increased margin requires using a product or service that is not the lowest cost, I'll use the more expensive product, even if that part of my cost structure is higher than it might theoretically be.This is not even to mention the topic of agility. For a CMO or LOB executive focused on making his or her numbers, getting to a solution and starting to earn margin more quickly is more important than getting the lowest possible cost for something. This is undoubtedly going to cause heartburn for IT executives planning private clouds because they can (putatively) be run less expensively than public cloud options. CMOs are focused on the here-and-now, not the someday-less-expensive.How-to: Track Cost Allocation for Cloud AppsLOB executives have to make their numbers. If they don't, they lose their jobs. No one will wait around for a supposedly better cloud service\u2014if executives don't start making their numbers right away, they won't be around to see the results.IT Must Understand Its Overall\u2014and Marginal\u2014CostIn a world focused on margin achievement, buyers are laser-focused on marginal cost, asking "How much does making the next unit cost me?" With public cloud computing, understanding unit cost is much more transparent than traditional IT cost assessment models. Moreover, the OpEx model is much friendlier to margin assessment than a CapEx-based approach.Consequently, for IT shops that seek to induce CMOs to place their applications into internal private clouds, offering fine-grained chargeback based on resource use is fundamental. Showback with deferred cost assessment is useless for CMOs or LOBs trying to assess a product's profitability. A different way of saying this is that CMOs will expect to be able to buy IT resource on a retail basis rather than being expected to fund a wholesale offering.This means IT organizations that want to be in the infrastructure game will have to provide fine-grained chargeback and stand ready to bill users on a pay-per-use basis. In addition, those users will expect to have the same flexibility with regard to use of the internal cloud as they have with external clouds. If the need for resource diminishes, users will expect to be able to release the resources back to the IT organization, and to incur no further charges.Commentary: Cloud Makes Capacity Planning Harder: 3 Fight-Back Tips and How IT Can Become a Cloud Service ProviderThe typically unseen challenge in this for IT organizations is that they will have to truly understand their cost structure and their marginal cost for delivering a specific resource\u20141 GB of SAN storage for a month, for example. In effect, IT organizations will have to act as their own CMOs, figuring out how to price their offering to cover costs, spread across a given user load.Managing utilization will be critical in this situation, as the marginal cost of a resource can vary wildly depending on how much the fixed cost associated with the resource can be allocated across a larger or smaller user population. One of my cloud computing predictions for 2013 is that IT organizations will need to understand their economics, and the move to CMO-led IT spending is only going to make that need even more important.If and when the shift in IT spending moves toward the CMO and fellow LOB execs, we can expect to see a significant shift in IT application profile and infrastructure spending patterns. We'll move from lumpy capital investment to pay-as-you-go OpEx, albeit with a new set of challenges associated with the new model. We can definitely expect to see much more focus on the business metrics on IT resource use, with an increasing emphasis on marginal cost and marginal revenue associated with granular IT resources.Bernard Golden is the vice president of Enterprise Solutions for enStratus Networks, a cloud management software company. He is the author of three books on virtualization and cloud computing, including Virtualization for Dummies. Follow Bernard Golden on Twitter @bernardgolden.Follow everything from CIO.com on Twitter @CIOonline, on Facebook, and on Google +.