How Advertising Budgets Are Like Fly Balls

New and better budgeting processes are a near-universal pursuit among businesses, particularly in advertising divisions. In an era of big data and powerful computing, the end of this pursuit is often assumed to include some type of algorithmic method — a mathematical technique that, once and for all, will identify the optimal amount to spend on advertising.

But what if there isn’t a single optimal tool? And what if experience and intuition can be as useful as data and algorithms in making budgetary decisions?

Generations of marketers and advertisers have been taught to view algorithmic budgeting techniques as more rational and therefore superior to heuristics — problem-solving mental shortcuts like rules of thumb, educated guesses or even intuition, which often lead to faster, more customizable decision-making. Consequently, managers have been reluctant to even admit to using heuristics in their budgeting decisions.

Our research addressed the question “What usefulness, if any, do heuristics have in determining advertising and promotion budgets?” In addition, we examined how prevalent these methods actually are — something few researchers have explored to date.

Prevalence of Heuristics in Advertising Budgeting

In our study, we sent a questionnaire to 1,000 advertising and promotion managers nationwide. We then evaluated the responses (125 total) with respect to four categories of information — principally, budgeting techniques used in their advertising budgeting process. Respondents chose one or more of 11 options, seven heuristic (e.g., advertising-to-sales ratios) and four algorithmic (e.g., ROI).            

Budgeting Method: Heuristic

• Arbitrary: Solely determined on the basis of what is “felt” to be necessary
• Affordable: Determine how much can be afforded
• Competitive Absolute: Set in line with the closest rival
• Competitive Relative: Set in line with market share
• Percentage of Last Year’s Sales: Might also be a fixed rate per case or non-dollar measure of sales
• Percentage of Anticipated Sales Next Year: Same as above, except uses forecast of sales to set budget
• Unit Sales: The company allocates a fixed percent of unit price for advertising and promotions and then multiplies this amount by projected sales volume (e.g., 5 percent unit price x 200,000 units sold)

Budgeting Method: Algorithmic

• Incremental Testing: The budget is allocated in an incremental series of market tests. Spending is increased or decreased in line with results.
• Objective Task: We start by setting particular advertising and promotions objectives and then derive a budget that will enable us to achieve these.
• Quantitative Models: Computer simulation models are used involving statistical techniques such as multiple regression analysis.
• Return on Investment (ROI): Advertising and promotions are considered investments and monies are spent to maximize ROI.

The results indicated heuristics are used surprisingly often in the advertising budget process. Nearly three-fourths (72 percent) of respondents said they rely on heuristics — 41 percent solely, 31 percent in combination with algorithmic methods. Furthermore, most companies use two or three methods, not just one.

Thus, although market data may be increasingly available, and sophisticated methods are placed regularly on the budgetary pedestal, our study shows most companies are not using these more analytical methods — at least not exclusively.

Budgetary Categories and Methods

Corporate Culture and Other Influences

In addition to reporting their budgeting techniques, respondents characterized their companies according to three “antecedents” that we suspected may drive the nature of the budgeting process — rooted in a theory about cognitive styles and their influence on decision making.

1. Corporate culture. We suspected some business environments would favor algorithmic budgetary methods, while others would tend toward heuristics. Our study confirmed that nearly all corporate cultures favor the use of a combination of heuristic and algorithmic budgeting techniques, except one — an adhocracy — which relied on algorithmic techniques alone. An adhocratic organization is marked by entrepreneurial spirit, emphasizing creativity and innovation. In these environments, power may be decentralized and corporate memory limited. Thus, these companies likely use more analytical methods to cross the “sufficiency threshold” in the budgeting process.
2. Risk-taking propensity. We hypothesized that a high-risk-taking company would use heuristic methods more often. This proved false; a company’s propensity to take risks did not seem related to its advertising budget techniques. Rather, an array of tools in budgeting seemed useful regardless of risk level.
3. Knowledge and experience. We looked at the relationship between the use of heuristics and an organization’s collective marketing knowledge and experience. As predicted, greater experience and knowledge is associated with greater use of heuristic techniques. In other words, over time, advertising personnel gain in confidence and experience, and are more likely to rely on intuition and good sense rather than solely on analysis and logic.

What the Findings Mean for You: Taking a Balanced Approach

In a classic 1951 Harvard Business Review article, Joel Dean noted that the fixed-percentage of sales method, a common budgeting heuristic, “gets the cart before the horse.” In his words, “Advertising outlays should cause sales, not be determined by them.” We contend that, although he may have been correct from a purely logical standpoint, this heuristic — and others — can be valuable nevertheless.

Specifically, our research concluded that heuristics used sensibly — in conjunction with algorithmic techniques — can serve as a vital check against data-intensive methods, plus save time and resources. Heuristics, by their very nature, are easy to use and adapt quickly, and are therefore appealing for very complex situations or when time pressure is high.

Consider the analogy to catching fly balls, from German psychologist Gerd Gigerenzer’s work on “fast and frugal” heuristics. A coach often instructs a baseball outfielder to “run to where the fly ball will land.” Yet, to catch the ball, fielders need not calculate the exact ball speed and trajectory. They simply begin running toward the ball, constantly adjusting their speed to maintain as constant an angle as possible.

Companies can think of heuristics in a similar way. For example, in contexts where advertising managers have accumulated good intuition and sense over time, heuristics may prove not only expedient but also often arrive at the same or similar outcome as algorithmic methods.

By the same token, formal analytics can serve as a valuable check, helping to ensure heuristics are used but not abused. In certain contexts, algorithmic techniques should be relied upon more — for instance, where company memory or managerial experience are lacking, such as a startup environment.

Transparency in the Process

A key conclusion of our research is that there exists a need for companies to be more open about their advertising and promotion budgeting techniques. Just as improvement in manufacturing processes began with articulation of those processes, advertising budget processes will only improve if managers become more willing to document them. Our research gives managers confidence that they can do so without fearing that they will appear irrational or lazy. Just because a technique is not scientifically rigorous does not mean it’s not sensible. In fact, data-intensive techniques can be as misleading and wrong as heuristic ones. Irrelevant data may be used naively when it should be forgotten, and the external environment may change so quickly so as to render past data impertinent. Furthermore, sometimes a manager may wisely apply a heuristic when cash flow or other pressing concerns clearly take precedence.

In summary, we suggest advertising managers proceed through the budgeting process heeding the famous advice, “It is better to be vaguely right than precisely wrong.”^[1] Use a blend of techniques, letting each balance the other, and stop looking for that elusive “best” algorithmic method. Recognize that sophisticated does not automatically mean better — and sometimes, it’s worse, entangling your process and wasting precious time and resources. As Oliver Wendell Holmes said, “I would not give a fig for the simplicity this side of complexity, but I would give my life for the simplicity on the other side of complexity.”

Above all, remain transparent about what actually works over time, since experience truly is the best teacher.

Paul W. Farris co-authored “How Corporate Cultures Drive Advertising and Promotion Budgets: Best Practices Combine Heuristics and Algorithmic Tools,” named Best Academic Paper by the Journal of Advertising Research, with colleagues Douglas West of King’s College London and John B. Ford of Old Dominion University.

[1] The quote is often attributed to economist John Maynard Keynes; however, Keynes’s wording was actually based on a very similar, older quote by logician and philosopher Carveth Read.