Everyone loves to talk about ROI. The benchmark has firmly planted itself in the soil of marketing doctrine, widely accepted as a measure that makes it simple to evaluate marketing programs and gauge spending levels. Return on investment enables financial types to evaluate marketing initiatives with the same approach they use to evaluate capital expenditures and acquisitions.

For senior management, ROI makes marketing less subjective. For marketers, ROI makes decisions fairly simple; it removes the appearance of subjectivity from the debate. Perhaps more important, a good return makes it easier to defend marketing initiatives and to justify existing and future spending.

But to us, there is a fundamental problem with overemphasizing ROI as the single measure of marketing success: It is often impossible to accurately quantify the impact. Although the world of marketing has come a long way in terms of analytic capabilities, applying financial numbers to the marketing equation is not always possible or preferable. That's why using ROI to evaluate the overall effectiveness can be a problem.

Take branding, for example. For many companies, brands are their most valuable assets. Determining the precise value of a brand at any given moment, however, is near impossible. Many companies do attempt to value brands. Interbrand and BusinessWeek, for example, produce a ranking of global brands each year based on brand value. Yet these brand valuation calculations generally rely on an assumption, built on an assumption, built on another assumption, built on, yes, another assumption. That means that trying to determine the impact of a TV spot on a brand's value or equity might produce numbers that are directionally right but certainly not precise.

Assuming brand impact
If the value of a brand cannot be precisely calculated, and thus known, then it would appear impossible to use solely ROI to evaluate the decisions that impact the brand. Either the impact on the brand has to be ignored, which seems incorrect, or it has to be put in as an assumption, which makes the analysis suspect.

That creates the potential for suboptimal decision making. Consider, for example, a company evaluating a cost-reduction idea. The capital cost is known, and the savings are known. The exact impact on the brand, however, cannot be determined quantitatively. Will anyone notice the change? Will it affect how they feel about the brand? How will sales and profit be affected? When will all this occur? All these questions can be hard to answer.

So on one side of the debate are certain, quantifiable savings. On the other side of the debate are qualitative concerns that the brand might suffer. A sharp finance executive could simply ask, "Well, what is the impact on the brand? Let's just factor that into the ROI calculation." The problem is that factoring this in with any precision is difficult, making the results of any analysis incomplete at best and misleading at worst.

Branding isn't the only area where returns are hard to evaluate. Customer satisfaction, loyalty, employee morale and differentiation are all important but generally impossible to put into hard financial numbers that can be used on a day-to-day basis when calculating ROI.

Super Bowl impact
Given these observations, it would seem that attempting to calculate the precise ROI on a Super Bowl spot, for example, would be a futile task. Certainly the cost is known, and it is rather substantial. But the returns are impossible to fully quantify. Consider Budweiser's recurring and considerable commitment to the Super Bowl. What is the net impact of Budweiser's appearance on consumers' affinity for the brand, the morale of employees, the motivation of distributors and the differentiation of the brand? While all of these elements are important to building a brand, they are generally impossible to put into hard figures that can be used to calculate an accurate amount.

Some marketing initiatives are easier to quantitatively measure than others. For example, a short-term price promotion generally has a substantial and immediate impact on sales. That makes calculating the return a fairly straightforward process. Similarly, online-advertising initiatives and coupons are often easier to evaluate with ROI. New-product initiatives may also be easier to evaluate using ROI, though as Clay Christensen, Stephen Kaufman and Willy Shih warn in their recent Harvard Business Review article, the calculation is often misused.

But here's the danger: A company that uses ROI to guide marketing decisions might focus only on initiatives that come with strong, quantifiable returns. The company might then reduce spending on programs that build the brand, increase customer loyalty and strengthen differentiation. That can be dangerous in that it increases the focus on short-term initiatives at the potential expense of more-valuable long-term gains.

'McStarbucks'
There's no clearer example of this than Starbucks. Over the past several years, the ubiquitous coffee chain has rolled out a series of initiatives to boost short-term profits at the risk of potentially damaging the brand. The ROI on each decision was probably very positive. But as CEO Howard Schultz admitted, the initiatives have hurt the brand and weakened the company overall. The new breakfast sandwiches, for example, might generate incremental revenue, but leave the stores smelling like cheese factories and make the baristas feel like they are working at McDonald's. As one rather frustrated barista noted on a recent visit, "Welcome to McStarbucks."

This is not to suggest that marketing should be a black box, where money is spent in a haphazard manner on programs that may or may not make any sense at all. Marketing decisions do need to be made strategically, and a marketing executive should manage and be evaluated on the overall financial performance of a business. But rather than focus on ROI, executives need to use a variety of measures to evaluate marketing programs' success. Those measures should be grounded in the objectives for a particular initiative. For example, if the goal is to strengthen customer loyalty, then loyalty should be measured and tracked. Anyone advertising in the Super Bowl needs to have clear objectives and a clear strategy, along with some way to evaluate success with respect to these strategies.

So what should managers do when asked to produce the ROI for a marketing initiative? Take a more open-minded approach to measurement, first focusing on a company's objectives and strategies and then identifying measures that can best work for them. Focusing solely on ROI is dangerous and na?ve.