Let’s face it: Data is a double-edged sword.
Customer data, in all its forms, is a treasure trove of information with which to improve brand performance. As enticing as it is, data also sparks a tremendous sense of anxiety — often overwhelming marketers with its seemingly endless volume, the uncertainty over whether to invest or retain ownership over first-party data and the pressure that having so much data at their fingertips puts on everyone from the CMO down to produce results that increase both efficiency and ROI.
Furthermore, there is a growing sense of concern in the industry that creative storytelling is getting lost in the brand conversation amid all of the intense focus on data.
While it is certainly true that marketing at its finest can motivate and inspire consumers to take action, the traditional marketing model is becoming increasingly outdated. Namely, it lacks a strong underpinning that seamlessly intertwines predictive data analytics and creative processes. Only when brands are able to consistently and reliably leverage data in the most effective ways will they be able to create more relevant advertising for the consumer. It is in this environment that we introduce our new model, called Marketing Engineering.
“Marketing today is clearly not reaching its full potential. In order for it to become mission critical for all companies, it needs to change drastically,” says Rick Milenthal, Chairman and CEO of The Shipyard. “CMOs are consistently being pressed to prove ROI on digital, eliminate inefficiencies and manage complexity. They need the data, the strategy and the ideas to navigate for success. Marketing Engineering is our approach to ensure their job is ‘future proof.’”
“Marketing today is clearly not reaching its full potential. In order for it to become mission critical for all companies, it needs to change drastically.”
Rick Milenthal, Chairman and CEO
Marketing Engineering is a new way of thinking about the marketing industry, its people and performance. It is a more thorough and accurate method of unlocking the true drivers of consumer behavior — a scientific approach to building marketing solutions in which the interplay of data and creativity drives better outcomes at each stage of campaign development. By harmonizing millions of data points in a structured way, we create a deep, personalized understanding of consumers to predict their behaviors and aspirations. We then deliver brand inspiration that is valuable, compelling and effective.
Marketing Engineering lives at the intersection of data, technology and creativity. It brings together two groups of people that have historically been sequestered: data scientists and creative professionals. Our agency is a fluid and interactive environment in which creative teams provide valuable input on data analysis (and vice versa) and where creative execution is stronger because it is grounded in deeper, data-driven insights. Often seen as disparate pursuits, data and creativity blend harmoniously in a Marketing Engineering approach that we affectionately refer to as “Data and Creativity Living In Sin.”
From an execution standpoint, Marketing Engineering has several major advantages, which are encapsulated by our “3Ps”:
Unlike segmentation models that reduce large portions of the population down to five or six roughly drawn personas, Marketing Engineering generates thousands of relevant entry points in order to significantly expand a brand’s universe of potential buyers. It allows us to capture all of the characteristics that define real people and real audiences. By bringing all of available data together in a structured way, we can determine which audiences will be most responsive to a particular product, message or media mix.
Marketers and agencies tend to think that the ultimate goal of prediction is more precision. We think of precision as table stakes. Brands succeed not when ads are targeted better or more precisely but when they make a personal connection; that’s the essence of personalization. Getting there requires the marketer to engage in a dialogue with millions of people in order to figure out which messages are actually connected to what consumers understand or are likely to motivate or persuade them.
Marketing Engineering includes a disciplined experimentation process that allows brands to know exactly what is working and what isn’t. A campaign may start out with hundreds of different that ultimately get culled down to 15 or 20 of the most effective ideas, using tools such as a Creative Optimization Engine (discussed below). Once you predict and personalize, you’re almost guaranteed to perform: Brands that personalize grow two to three times faster than those that don’t.
Marketing Engineering cuts across all aspects of building solutions and can significantly improve performance in each of the following areas:
Marketing Engineering can be applied in any situation that requires a brand to test new ways of thinking or new value propositions. This includes not only standard solutions (new product launches, re-branding campaigns, etc.) but also strategic business initiatives such as when a company wants to introduce a new trade channel (e.g., direct to consumer). In cases where brands need to spend less or spend differently, the discipline of marketing Engineering can get them there faster than basing decisions on instincts or cutting the parts of the budget that are easiest to cut.
“This is a completely different way to extract rich, real-life insights,” says Dave Sonderman, Chief Creative Officer of The Shipyard. “Using advanced data techniques to find out what a million customers are doing or what they believe in is much more actionable than dissecting the comments of twelve people in a focus group.”
Any “engineered” marketing solution begins with a comprehensive approach to collecting and analyzing all of the different classes of available data, including first-party purchase/POS data as well as various categories of behavioral data (both online and offline) from hundreds of independent third-party providers. Various customized features that leverage digital technology and data science then begin to build the solution.
“Leveraging data within a marketing Engineering framework is about understanding all of the information you know about a customer that could provide a competitive advantage, and then making that data available in a format that can be consumed programmatically or at least analyzed in real time,” says Ben Clarke, Co-Founder of The Shipyard. “This often entails making the data streamable, instead of just reporting, and working with publisher partners, IT and others to be able to consume the raw feeds.”
One tool that can significantly improve audience identification and data aggregation functionality is a custom-built data management platform. For example, let’s say that a consumer products marketer wants to define a brand’s target audience within the total U.S. female population using five major attributes (e.g., yoga enthusiast, Whole Foods shopper, Subaru buyer). There could be as many as 30 to 40 audiences that fit each of these descriptions—so how does the marketer know how many and which ones to choose from?
“A standard DMP does not provide a systematic way to identify those audiences or to roll them up based on those descriptors,” explains Joel Acheson, Chief Marketing Technology Officer of The Shipyard. “Our data back-end is connected to 40,000 third-party audiences. We use a text mining application to identify composite audiences that are most likely to work and roll those IDs up through an ad server versus pushing them out manually one at a time. This allows us to be smarter and more efficient about who we are targeting, and to do it in a way that is systematic and repeatable.”
Advanced techniques such as creative experimentation may also be used to enhance audience identification and test campaign messages in many different permutations. This process leverages artificial intelligence, machine learning and social listening tools to identify all possible audiences that—when paired with the right messages—are the mostly likely to drive awareness, trial and conversion.
In the case of that same CPG marketer, there may be at least four or five distinct categories of audiences/messages that are used in different combinations depending on the campaign strategy or tactic. A representative sample might include educational messages, product benefits, ingredient information, corporate responsibility and brand recognition. Each of these sets of attributes may appeal to consumers in varying degrees in different contexts, but ultimately all are highly predictive of a desired behavior or outcome.
Marketing Engineering is a process of continuous learning in which data science and creative teams work hand in hand to update and optimize the approach. Traditional research is used a starting point. For instance, if a brand or retailer has established a particular segment, teams can go back into the DMP to try to replicate that identity and build on it.
The steps leading up to establishing a creative platform blend audience identification with message testing to create a series of hypotheses about campaign performance. Potentially hundreds of campaign ideas may be tested for a particular response (i.e., click on an ad, visit a website) and the results compared to insights derived in the initial non-targeting stages from various social listening techniques in order to determine the optimal collection of target audiences and supporting positions.
“The creative gets better and the messaging is more personalized as the process evolves because we spend more time on each ad,” notes Porigow. “It’s a cumulative process of building on learnings from previous tests. We’re crafting each piece with more knowledge of what people are responding to and gaining a better understanding of what to put back into the marketplace.”
As these steps are repeated throughout the campaign process, it allows the marketer to build a creative experimentation engine that can be used to refine and optimize everything from prospecting and remarketing to creative and copy testing.
All of these tools and methods help build a knowledge base that goes far beyond the scope of an existing campaign. Through the process of learning what messages are working and adjusting in real time, brands can not only optimize media delivery but also begin to anticipate trends and respond in impactful ways.
Such learning may afford brands a tremendous competitive advantage. In the current state of marketing, most brands are working off the same sets of data and insights, or at least have access to the same information. What separates one brand’s performance from the next comes down to creative interpretation—i.e., what they do with the data. Imagine what would happen, however, if suddenly all brands were not working from the same place.
“We see a future in which brands are separated from one another by the information they possess about their consumers,” says Dave Grzelak, Chief Strategy Officer of The Shipyard. “We’d position this to marketers by saying, ‘What you know about your customers is unique to you, and that provides you a different lens and a way to go to market that truly is about who your customers are as individuals and what they respond to from a marketing perspective.’”
Adds Milenthal: “Marketing engineering is about empowering CMOs, understanding their fears and giving them a message that matters. Our approach is very calming and helps them better understand the data and the process that allows you to get to the end result, which is higher performance marketing and personalization. It is a way of really owning and connecting all of the different sets of data in a way that gives marketers a unique view of their customers and the power to move these customers to action.”
The time for CMOs to consider a Marketing Engineering approach is now. It is a critical step to future-proof their roles, because the speed of change brought about by new technologies in the marketing sector is only going to get faster. Disruption comes fast and furious. We’ve got to be ready.