Forget Mad Men. Today’s marketers are more likely to be math men and women. They plumb the depths of Big Data with advanced analytical tools. They buy and use dazzling new software—some spend more on technology in a given year than their companies’ IT departments. They are hot on the trail of marketing’s holy grail, the ability to measure return on investment (ROI) on every campaign.

These are laudable pursuits, up to a point. The trouble comes when marketers mistake the algorithm for the person. Humans, after all, are pesky creatures. We make decisions with our hearts as well as our heads. Our wants and needs can’t always be reduced to, or predicted by, data. Great marketers have always known this, which is why they work hard to build an emotional bond with the customers they are targeting. They tell compelling stories about their brands through memorable messages and indelible images. At its best, this kind of marketing pops and dazzles, like magic. Marketing that ignores the magic and relies on math and science alone will be marketing that doesn’t work.

So there’s a challenge here. Today’s marketers can’t succeed without the new technological and analytic tools; the tools offer insights that simply weren’t available in the past. But marketers also need a holistic view of the people they are trying to reach, in order to forge emotional connections. The two approaches—math and magic, to use shorthand—involve different skill sets and ways of thinking. Can they be combined?

We think they can, because a handful of leading companies are already assembling the ingredients of just such a combination. These companies focus relentlessly on their marketing priorities and their target customers, never getting sidetracked. To implement the priorities, however, they are heading down new paths. They figure out how to integrate the torrents of data into a single overarching view of the customer. They view marketing not through the usual categories—channels, media and so on—but through the omnichannel perspective of consumers. Instead of relying on what happened last year, they run sophisticated tests to determine what will be most effective in the future. (After all, what works with one person won’t necessarily work with another.) They continue to tell stories about their brand and to connect with consumers’ hearts as well as heads, but they do so in new ways. Contrary to the inclination of many marketers, moreover, they do not put measurement of ROI at the top of their agenda. When they can’t measure the ROI of an innovative effort, they are willing to let instinct guide their efforts—at least for a while.

Infographic: Secrets of Marketing Leaders

In a survey of marketers at 436 companies in the US and Canada, we compared the leaders—those who were increasing sales and market share—to the laggards, who were losing sales and share. The leaders were three times as likely to use data in their decisions. And they were twice as likely to own their own creative capabilities, the ultimate source of great stories and memorable messages. Let’s look more closely at the central elements of their math-plus-magic approach, and how these marketers use it to stay connected to customers.

1. A single smart view of the customer

Most major companies these days have access to large quantities of information about their customers. Marketers can capture and analyze sales patterns in ever-greater detail. They can track someone’s locations and transaction histories. They can draw on new sources of information, such as what people are saying about the company on social media. Because data flows into a company in real time, they can see what their customers are up to right now. In nearly every case, however, this data sits in different places—a company’s legacy CRM systems, its loyalty program databases, new cloud-based applications, third-party servers—and can be hard to access. Sometimes marketers must wait in line to get the data they need served up to them by IT.

Suppose that all data could be integrated into a single overall view of each customer. Suppose the records could be updated immediately, with every additional transaction or piece of information instantly accessible to anyone using the system. A company would get a remarkably full and intelligent view of its customers. When a customer calls a company, call-center reps can see what she called about before. But the rep won’t know what she has purchased through other channels, whether she has returned merchandise to a store or what she has said about the company on social media. If the rep could see all that, it would make for a much more productive service engagement.

The first step toward a single smart view is always organizational, and involves ending the traditional turf wars between marketing, IT and other functions. People from each group need to come together to share priorities and develop an overall plan. Some companies have created the new position of chief marketing technology officer to oversee the collaboration. Others stick with the existing org chart but make changes to their operating model so that the functions work together rather than at cross purposes. They typically expect IT to become more of a business partner, developing a joint roadmap and investment plan with marketing as part of the annual resource planning cycle. They then hold the two functions jointly accountable for delivering on the plan.

The next step is creating the new integrated customer records, a process that generally goes by the name of master data management (MDM). An MDM system sits on top of existing databases, capturing everything from store traffic patterns and online behavior to comments on social media. Every interaction enters the system in real time, and users of the system can tap into it from wherever they happen to be.

While no company has developed a complete MDM system, some are getting closer. Consider Hilton Hotels’ extensive use of data. After a guest stays at a Hilton in Houston on business, he browses the Hilton website to explore a week-long vacation stay at a Hilton in Paris. When he calls central reservations with questions, the call center representative will thank him for the recent stay and recognize that he was reviewing Paris hotels on the website. Hilton’s automated system connects the guest stay history, web browsing and call data, third-party information that the guest splits spending between Hilton and Starwood, and projected occupancy data for Paris properties. Data management thus allows Hilton to improve its cross-selling, balance occupancy across properties and make the process faster and more convenient for the guest.

None of this is easy. Companies need to crawl before they can run, starting small and building from there. The objective is to discover previously unknown insights, and that often requires many hit-or-miss attempts. Moreover, it’s essential not to let the perfect be the enemy of the good. Companies often have hundreds or even thousands of databases, some of them built many years ago. Integrating them all into a single MDM solution tends to be cost prohibitive. Instead, savvy marketers focus on the handful that matter most and concentrate their resources there.

Liquor company Rémy Cointreau’s marketers, for example, didn’t wait for a full-scale MDM system. Instead, they decided to act on what they could access themselves. Social media data helped them know not only what people were saying but where they were imbibing—and that, in turn, helped the company target brands to particular bars with better accuracy.

インフォグラフィック

Secrets of Marketing Leaders

How leading marketers combine math and magic to stay connected to their customers.

2. Digical® experiences—and Digical marketing

Companies now provide all sorts of what Bain calls Digical experiences—products or services that combine digital and physical elements and thereby create new sources of value for the business. Google’s Internet-connected Nest thermostat and similar devices reshape the experience of maintaining a home. Disney’s MyMagic+ system transforms the experience of visiting one of the company’s theme parks. Retail shopping, in particular, is undergoing a Digical transformation, as consumers move seamlessly from online shopping to physical shopping and back again.

Digical marketers use new sources of information to learn what annoys customers and what delights them. Building on their unified data files, they create consistent messages and experiences for customers across every platform. They use digital data to enhance the physical experience and vice versa, and they adapt quickly to changes in customer behaviors and preferences. Of course, they also augment the master file continuously with each new customer interaction.

The result is a kind of digital magic, because experiences like these make customers feel that the company knows what they want and helps them find it. Sometimes the magic is simple. The grocery chain Wegmans, for instance, created a mobile app that lets consumers select a recipe at home. The app then adds the ingredients to the customer’s shopping list and provides a route through the store layout.

Other companies have developed more complex digital-physical magic. The Chinese cosmetics company Jahwa provides reps on the sales floor with mobile devices that can tap into an integrated database. When the rep puts in the name of the customer she’s serving, the device provides data and recommendations tailored specifically to that customer—similar to the recommendation engine that consumers now experience on many digital sites.

This logic applies in business-to-business markets as well. A large machinery manufacturer automatically collects electronic data about end users and product usage. That lets the company’s marketers understand the needs of highly specific groups of buyers and design appropriate offers.

Virtually every product or service these days, from hotel stays to earth-moving machinery, combines digital and physical elements, although Digical innovations will hit some industries harder and faster than others. That’s critical to the new math-and-magic marketing because it allows marketers to not just create seamless, personalized experiences but also to do what digital-native companies have always done: test, learn, adapt and test again.

3. Relentless testing and learning

Digital marketers at companies such as Netflix and Amazon constantly experiment with offers, pricing strategies, website design elements and countless other variables. Continuous testing infuses marketing with a dynamic element, allowing marketers to learn more every day about their customers’ priorities and behavior. It also creates an environment in which experimentation and innovation are the norm. A smart view of the customer augmented and refreshed by Digical experiences enables physical-world marketers to do nearly as much testing and learning as their online counterparts. Instead of depending on historical data, they can create future-oriented hypotheses. They no longer have to bet heavily on one set of messages because they are essentially transforming traditional one-way marketing communication with customers into two-way communication. When they listen to what customers tell them, they can respond accordingly.

Multivariate testing, moreover, extends the power of experimentation beyond the traditional A/B “champion and challenger” test. This technique, called experimental design, allows marketers to test dozens or even hundreds of combinations of offers, messages and products. It can be used in direct-mail and email campaigns, advertising and call-center, and customer-experience programs (see Figure 1). It typically improves both response rates and conversions, and it speeds learning by testing many variables simultaneously.

A major US communications provider, for example, tested 17 variables (offers, messages, calls to action and so on) in 12 different combinations (see Figure 2). The best of these returned five times the response rate and from two to three times the sales conversion rates of the company’s top existing offers. The marketing team was able to track results beginning just a few days after launch, and it could monitor response fatigue. Using this real-time feedback, it continually refreshed the messages and incentives to lift response rates.

Intensive testing and learning can put a strain on companies that don’t prepare for it through extensive collaboration among functions. It may require more training and support, so that operations and customer service reps can manage a wider variety of interactions. And it requires a company to act quickly to capitalize on the flow of new information. Companies should start small and work up, gaining experience as they go. Clear financial thresholds such as profitability targets can speed up decision making and enable marketers to develop a repeatable test-and-learn model over time.

4. Realism about ROI

Today’s math-driven environment has awakened marketers’ inner CFO. At last, the thinking runs, we can finally track the ROI of every initiative and campaign. We will know exactly what the payoff of our efforts will be. Real CFOs echo this belief, and pressure marketers to gauge the financial effects of their every move, especially as the scale of spending on marketing technology grows sharply. Trouble is, the marketing world itself is changing rapidly and many new opportunities don’t lend themselves to easy ROI calculations. The leaders, we have found, recognize that it is just as important to experiment as it is to hold themselves to strict ROI requirements.

So what do these marketing leaders do? How can they justify spending on campaigns and channels where they can’t measure the impact?

Let’s take an example from social media. In Bain’s 2014 survey of 609 marketers who were spending on social media, we found that for those we identified as “dabblers”—companies spending less than 20% of their digital budget on social media advertising—ROI was the most important criteria for deciding where to spend (see Figure 3). As one respondent said, “Data, numbers and ROI are not nearly as possible for us yet when it comes to social media—but a big part of my job is getting us to a place where we can measure everything.”

Marketers who had moved past the testing phase and were spending more than 20% of their digital budget on paid ads in social media—the “believers”—took a different approach. Effectiveness of communication was still the most important criteria for this group, but they assessed effectiveness differently, by volume of leads and actions or reach of a targeted audience. One of these marketers told us, “The goal is to use all the different types of digital and social media to complement each other, but that’s only really possible if you have a great story to tell.... You want a medium that is effective for communicating that story to the right and, hopefully, large audience. Getting hung up on ROI doesn’t allow us to tell that story.”

In fact, the companies in this second group scored ROI next to last in importance. While these marketers also wanted to see the data and results that gave them confidence that their spending was productive, they were less focused on the calculation of true ROI. Looking at the full picture, including metrics other than ROI, made them comfortable with increased spending on emerging platforms. Testing marketers, stuck in the data, tended to keep spending the same.

“Getting beyond ROI,” as it’s commonly known, remains a critical part of effective marketing. Most marketers can’t know all of the actions that customers take on their journey through a company’s marketing. Even direct-response marketers don’t know for certain whether the response to an ad stemmed from the ad itself or from the fact that a friend happened to recommend the product earlier that day. While we understand the pressures on marketers—increased demand for spending efficiency, growing tech budgets and an endless supply of data to analyze—the most effective marketers complement their quest for ROI with a similar commitment to marketing magic.

Three things you can start tomorrow

Marketing has become complex, sometimes overwhelming. Marketers must find their way through many new technologies and opportunities. But the leaders seem to maintain level heads in the midst of all this activity. They use the new technologies, and indeed even expand on them, through MDM and other methods. They capitalize on the opportunities provided by Digical products and services, such as the ability to test and learn quickly and comprehensively. And they avoid a rigid focus on ROI lest it hold them back. Distilling their approach into three things that other companies can act on, the list might look like the following:

1. Paint a holistic picture of the customer. Data analytics are powerful and can generate highly useful insights. But bits of data alone do not describe the people who are customers. Work on integrating your data so that you can assemble it into a more complete portrait. Supplement the data with human interaction, including direct customer feedback.
2. Encourage emotions. Marketing has become a two-way street, with customers talking back to companies through Facebook, Twitter and many other channels. Much of this communication will be emotional. Do they love you? Hate you? Why? These are valuable insights. But you can’t respond to emotions purely with tweaks to your offers—you need some magic, a mix of the unexpected with the occasional “Wow!” It isn’t enough to rely on the math.
3. Stay nimble—and be bold. In fast-changing markets, you can be the first to try something new. The “learn” in “test and learn” means being able to discard what isn’t working and devise more effective efforts. New opportunities are often untried, and only the fast and fearless will be able to take advantage of these opportunities.

The marketing world of Mad Men was entertaining, but it’s gone forever. The new world of math-and-magic marketing is at least as interesting—and is proving to be a great deal more effective.

Laura Beaudin is a partner in Bain & Company’s Customer Strategy & Marketing practice, leading the practice’s Marketing Excellence work. She is based in San Francisco. Mark Brinda is a partner in Bain’s Technology, Media & Telecommunications practice, based in New York. Jason Ding is a partner in the Customer Strategy & Marketing practice, based in Beijing.