The following is the 1st installment in a six-part series examining best practices for the measurement of shopper marketing. This article looks at effective ways to rationalize the investment. Subsequent articles will cover the measurement of shopper behavior and brand impact, effective integration practices, retail collaboration and directions for the future.
PART 1: Rationalizing the Investment
Shopper marketing promises big things. It promises to fuel brand demand and influence shopper behavior while also selling more product in-store. Brand teams like it so much that, according to Shopper Marketing’s annual Trends Survey, over half plan on increasing their shopper marketing budgets. Some of this increased funding will be siphoned away from above-theline efforts. Yet, despite the growing enthusiasm for t his approach, most brands still don’t understand exactly what they are getting for their money.
When the basic unit of success is a dollar, a shopper marketing program’s other achievements can be undervalued or forgotten entirely. Avoid the “measurement trap” wherever possible by tracking metrics other than sales.
The best indicator of success is what everyone agrees it should be. Getting agreement on success measurements upfront is crucial to achieving goals and building benchmarks.
The level of spending on a program should dictate the scope of most measurement plans. The greater the spend, the more sophisticated the metrics.
In general, there are three tiers of measurement by spend level:
Basic sales Shopper behavior Brand beliefs and attitudes
The bigger and more extensively measured programs can serve as tutorials for the rest of the organization, especially when systematically shared before the next planning cycle.
Avoid the impulse to compare and despair: the entire industry is on a learning curve. Keep learning and improving.
of foreign brands and in some cases are more trusting of foreign/global brands. In China, nine out of 10 consumers say Shopper marketing has been likened to quantum mechanics, the world of innitesimally small things. It is like micro-marketing on a mass scale. On the other hand, traditional marketing is like Newtonian physics, which takes a broader perspective on the world. Given these differences, older research methodologies (especially broad-scale, national ones) don’t capture the entire impact of smallerpenetration shopper programs.
Furthermore, shopper marketing often demands a broader, brand-agnostic mindset that requires different success measures beyond number of cases sold – brand-portfolio lift, out-of-store behavior changes, category lift and trip frequency, among others – that traditionally haven’t been part of the arsenal. That makes comparing the financial returns of shopper marketing efforts to those from traditional activity very difcult.
Throughout the industry, marketers adopting a shopper strategy have been grappling with these questions, “How do I rationalize my spending, above and below the line? What am I getting for my investment – really? What should be the key metrics across programs?”
To understand these issues and their potential solutions in greater depth, Shopper Marketing interviewed dozens of executives across the industry, from retailers to researchers, agencies to brands. The result is a series of six articles reporting on best practices in measurement and the direction research may be heading with advances in technology.
Key Issues and Hurdles
Shopper marketers are looking to rationalize investments as budgets shift from above to below the line. For most brands, there is a single pot of money for the totality of marketing efforts; it is a zerosum game at the beginning of each ning cycle. But a portion of the money that used to be spent gaining impressions through mass media advertising is now spent on shopper initiatives. The question for many is, “How do I justify this spend shift?”
Because shopper marketing is more closely aligned with the purchase than traditional media, it stands to reason that measuring its impact should be easy. Or at least easier. But this hasn’t been the case. Shopper marketers have been stumbling over various obstacles in their quest to measure impact. There seem to be four major stumbling blocks, recurring across categories and channels.
1. The Measurement Trap
Thirty years ago, promotions were one-off events that were fairly simple to measure. A brand would need to move a certain number of cases to pay off an investment in a dened period. This concept was easy to grasp. The measurement for success was one-dimensional: sales volume. But if you look at measurement through the shopper-marketing lens, it becomes more complex, because shopper marketing is an effort to provide long-term relevance by delivering holistic solutions to shopper needs – thereby building the brand and changing behavior in the process.
So what is the “measurement trap”? Defaulting to sales metrics, to the exclusion of other measures. Tracey Doucette, senior vice president, customer strategy, eld and shopper marketing, at PepsiCo, warns, “It is tempting to use the metric we can easily get, rather than measure the ‘real’ objective of the program.” But the best shopper marketing programs have objectives that extend well beyond sales into shopper behavior and attitudes. If only sales are measured, the full yield of shopper programming is obscured.
2.What Are We Measuring Anyway?
In an attempt to capture a fuller picture of program performance, other acronymed success measures have emerged, including: Return on Objectives (ROO), Return on Marketing Objectives (ROMO), and Return on Relationship (ROR), among others. ROO or ROMO usually refers to measuring shifts in shopper behavior over time (beyond the promotional period), or to shifts in brand attitudes. These changes can be determined quantitatively, albeit at signicant cost. The ROR is not often a quantitative measure, but a concept that attempts to capture the positive impact of a shopper program on the relationship between a manufacturer and a retailer.
However signicant these achievements may be, they can be undervalued or forgotten entirely when the unit of success is a dollar, and only a dollar.
Effective programs seek to change behavior well beyond the promotional period, exerting a lasting inuence on shopper habits. For example, the Kraft iFood Assistant mobile app suggests meal solutions that bundle products to solve a shopper’s “What’s for dinner?” dilemma. The app is also geared to meet business objectives: driving sales for Kraft products, and increasing basket ring for retailers. [Note: While some industry professionals restrict “shopper marketing” to activity related to particular store environments (either literally or through collaborative out-of-store programs), general consensus denes it as any activity that pushes a shopper along the path to purchase, the interpretation that will be used in this series.] The industry is seeking to understand how these kinds of shifts in shopper behavior can be captured in terms of metrics, and then compared with results from other programs.
Shopper marketing can also influence brand beliefs and attitudes. Think of the positive impact of the pink ribbon on Yoplait’s brand perception, or the halo that Campbell Soup’s Labels for Education program gives to participating brands. In addition to driving sales, these programs leave lasting impressions on brand equity. (This can hold true for both the brand and its retail partners.) But how is the brand equity impact captured here? And how do we gauge any effect on the manufacturerretailer relationship?
To add yet another layer of complexity, stakeholders have differing agendas and denitions of success. Predetermined metrics need to reect the objectives of the program, especially when multiple objectives are in play. Without exception, everyone interviewed for this series (agency, researcher, brand and retailer) asserted that stakeholders need to agree on the markers of success, from the outset, in order for success to be achieved. Programs are so diverse that it can become challenging to compare one to another in “apples to apples” fashion. Therefore, it is imperative to reach internal consensus on success metrics prior to program execution.
The chart on page xx identies three buckets of shopper-centric measurement: sales transaction data, behaviors, and brand attitudes/beliefs. Manufacturers and retailers each have goals in these buckets, but they can differ. Incorporating both sets of goals into programming has become critical to success. Deborah Hannah, shopper marketing director at Starbucks Coffee Co., says, “A self-serving brand goal doesn’t cut it in shopper – the shopper objective needs to incorporate retailer objectives, such as share of wallet, margin growth, category growth, and basket size.” (The task of reconciling objectives and managing the analytics process with retail partners will be addressed in article ve, Collaborating with Retailers.)
Collaborating with the retailer on objectives can bring its own rewards. The chart also includes a fourth bucket, the program’s impact on the brand-retailer relationship. While this is rarely a “hard” metric, each stakeholder hopes to receive some benet, or return, in this area as well.
3. The Dubious Protability of Measurement
Increasingly sophisticated marketing strategies have demanded increasingly sophisticated metrics. And, for the most part, these are available. From in-aisle video monitoring, to tracking in-store trafc patterns and eye movements, to linking online exposures to bricks-andmortar purchases, shoppers are being tracked, tagged, asked and observed. Methodologies abound. Why don’t results?
The answer is that funding state-of-theart metrics tools would wipe out protability for many smaller programs. It doesn’t make too much sense to spend $40,000 to measure a $150,000 program.
As a practical matter, then, not all shopper marketing programs can – or need to be – measured to the fullest. According to industry professionals from both sides of the table, it seems that best practices are tiers of measurement commensurate with spending. That is, the bigger the budget, the more extensive the performance metrics. The size of the program sets the level of expectation for the metrics; this makes sense from protability perspective, because programseffectively “buy” their own metrics.
Furthermore, the bigger and more extensively measured programs can serve as tutorials for the rest of the organization, especially when these are systematically shared before the next planning cycle.
Sticking with the three buckets as a simple way of parsing metrics, the scope of the program roughly parallels the metrics to be obtained (see chart on page xx). Virtually all programs, even one-off tactical efforts, are examined from a simple lift standpoint using sales data. Lift metrics are then compared to historical norms and benchmarks for that category.
A few larger, forward - leaning manufacturers have improved upon simple lift analysis to employ sales data as inputs into custom marketing mix models. (“Marketing mix modeling” refers to the use of multivariate regression and other statistical techniques on sales data to gauge the impact of various marketing tactics.) While nearly all of the larger consumer packaged goods brands run analyses annually, using national data as inputs, the more progressive brands are using desktop models on a brand-bybrand, program-by-program basis.
Programs with greater scope and spending may warrant obtaining the next level of measurement: shopper behavior information. This can come from shopper card data, when examined longitudinally and across baskets. Other kinds of behavioral metrics can come from research specically set up for the task, including shopping cart trackers, in-aisle video monitoring, eye-tracking, virtual store testing, and others. (Article two will discuss a number of these methods.)
The biggest campaigns, of course, are most likely to earn the most extensive analysis, which would include brand attitude measurement. While most brands examine shopper behavior (via panel data) and brand equity (attitudinal tracking studies) on a national basis, these metrics are not usually broken out by retailer or by program. However, the largest shopper marketing programs have the funds required to acquire these data on a oneoff basis.
4. Human Nature
Nearly without exception, the executives interviewed for this series were a bit apologetic about their self-perceived lack of rigor in measuring performance. Within the walled gardens of their companies, shopper marketers lament the shortcomings of their methods, even when they are relatively minor. These shortcomings seem more acute when combined with the sneaking suspicion that “someone” out there is doing a better job. It’s human nature to put one’s expectations a little above reality and adopt the “grass is always greener” method of comparison. But psychologist Robert Bringle was probably right when he described envy as
a positive motivator that inspires people to work harder.
It is not surprising that the marketing industry expects great things from measurement today. It has only been about a dozen years since the smartphone was introduced. Most professionals remember a time before smartphones, a time before cellphones and, for some, even a time before color TV. Technology is accelerating our capabilities and our expectations, too. Marketers now feel a compulsion to track every brand interaction with every shopper – accurately, cheaply and in real time. And this seems completely plausible. In fact, something similar to this scenario may be in the ofng through new technology.
In the meantime, however, it can be disappointing to be confronted with the realities of legacy measurement systems. But is this disappointment really warranted? Traditional above-the- line advertising never really delivered – or even promised – a direct return on investment. While relationships can be drawn, no metric has ever proved it. Shopper marketing is being held to a higher standard. Perhaps this is because new technology now tantalizes with the promise of perfect data.
“Measurement has always been a challenge. Traditional marketers, especially in brand advertising, have had to rely on attitudinal shift metrics, as there are many other inuences that occur between the brand advertisement and the purchase,” says Fred Bidwell, executive chairman of JWT Action.
“With shopper marketing, we have more inuence throughout the purchase funnel. While challenges still exist, new measures allow us to look at the full spectrum of attitudinal and behavioral shifts – including sales.”
New metrics for success are being discovered as new paths to purchase are being explored. The journey may be confounding at times, but the outcome – more effective measurement – is worthwhile.
Tracey Doucette Sr. VP Customer Strategy, Field & Shopper Marketing, PepsiCo
Deborah Hannah Shopper Marketing Director Starbucks Coffee Co.
Fred Bidwell Executive Chairman JWT Action