PROPOSITIONS

Big data is rapidly transforming a number of business functions across many industries. The biggest shift is in how we market to our customers. Unlike yesterday’s environment, where marketers broadcasted across a set of customer segments, we can now personalize the communications to each customer based on his/her current predisposition to the products being sold. I have been observing the growth of big data and advanced analytics and its impact on marketing analytics. I would like to make three propositions, which summarize my understanding of how big data and advanced analytics will change marketing analytics. These propositions were inspired by a talk IBM CEO Ginni Rometty gave to CMO community5 and have evolved as I researched the related topics and discussed with a number of practitioners.

Proposition 1: From “Sample recalls” to “Observingthe Population”: For many years, marketers sought small samples from their customers, used interviews to recall past behaviors, and extrapolated to reconstruct the buying behavior for the larger population. Marketers learned to create a number of sophisticated techniques for projecting results to the larger population using the smallest samples they could afford, as it was cost prohibitive to collect and store large data. Also, marketers had to conduct interviews in limited amount of time, leading to broad questions and inaccurate results. Marketers did not have the luxury of asking detailed questions, like how many products did you browse before making the purchase selection, and whether they were all similar or very different. Big data has changed the game completely. We can connect with customers, record every click on the Web, watch every step in the store, and listen to all the public conversations. Big data brings many more observations to the table, thereby leading to a lot more detailed analysis and discovery. Marketers now have access to a excruciatingly detailed understanding of shopping behavior. In addition, marketers are no longer dealing with small samples as we now have access to customer data for a very large proportion of the overall population. How does it change our understanding of the consumer? We need an improved vocabulary and a new way of interpreting statistical data. Let me use a sports analogy to illustrate this proposition. When I was about 12 years old, my father first showed up to watch our cricket game. While he was trying to understand the game, he became far more fascinated with the scorekeeping process. “I wish they would do this level of detailed scorekeeping at work, and we would have a lot more objective performance evaluation,” he told me as he watched us painstakingly record every ball and score. Somehow, most sports, such as American football, soccer, tennis, or cricket, carry their own versions of very well-defined methods for creating a great number of detailed observations. Imagine an analysis of American football without passing yards, touchdown passes, and interceptions. Without this statistic, we could still have the resulting scores, but would not be able to deduce the quality of the game. A sample recall is like knowing the final score, while a detailed observation is like knowing the passing yards and the interception information that contributed to the final scores. Imagine a set of researchers standing outside stadiums asking the audience leaving a game how they would differentiate the winning team in the absence of these observed statistics. The observations have provided a vocabulary and a way of comparing players and teams, and this information has been available long before automation brought us closer to big data in other areas. As automation goes up and storage cost goes down, marketers are seeking an exponential rise in observations in many areas, creating a new set of metrics to define and measure customer needs and interests. Product automation has enabled corporations to collect a lot more data, and corporate IT organizations are getting better at storing rather than deleting all the event-level details.

Proposition 2: Marketing through collaborative Influence: Marketers are seeing a rapid rise in collaborative influence from marketers to customers, from customers to marketers, and from customers to other customers, each of which can be fine-tuned to a large number of microsegments and used for personalized communications. Marketers can converse with the customers as they make decisions, and influence their decision-making using a series of sophisticated marketing tools. Customers can influence how new products are developed and how existing products are evaluated. Marketing in the broadcast era was all about repeating the same message over and over, until it reached desired reach and opportunity to see (OTS). Elias St. Elmo Lewis, cofounder of the Association of National Advertisers and one of the first advocates in the field of advertising, created the AIDA model (awareness, interest, desire, action) around the beginning of the twentieth century, which described how marketers influence customers.6 In the early days, it was not possible to observe who saw the communication and how they reacted to it. Today’s collaborative influence is far more of a two-way persuasion process. Analytics allows marketers to identify customer needs, initiate a dialogue, and customize a collaborative process to converse with customers, whether with the help of experts, targeted advertising, or through personalized communications directly with the customers. Also, crowdsourcing is increasingly used for product idea generation or evaluation. Many product ideas are based on ideas or feedback from the customers. It is possible to listen to customers and act on the communication received.

Proposition 3: From silo’ed to orchestrated marketing. Various marketing departments, such as marketing research, advertising, product management, and sales, had limited avenues for orchestration. We did not know that a specific customer was interested in purchasing a product, and had been to various websites, and hence was a qualified prospect for additional information or promotion. So, the marketing effort could not be targeted and orchestrated across the departments. Using big data and collaborative influence, however, the market leaders are beginning to orchestrate their marketing investment, and focusing their attention on their customers in ways we have never seen before. The four P’s—product, pricing, place, and promotion—are no longer offered via a set of silos broadcasting their overlapping messages. A set of sophisticated orchestration engines takes into account customer privacy preferences, needs, intentions, and the current relationship with the marketer to coordinate a set of actions. It is like having an episode of a television sitcom that can be changed based on viewer preferences. The observations are generated using a series of sources, most of them external to the organization. The messaging is delivered via a complex web of marketing organizations, each specializing in a specific instrument. However, the marketer keeps track of customer status and directs the messaging accordingly. If the customer is searching for a product, he/she may at his/her own initiative starting pulling the information. As marketers discover an unmet customer need, they are able to respond back with marketing information, up to and not exceeding a saturation point. If I am shopping for a smartphone, the minute I start searching the Web, tweeting to my community, or shopping at stores, the marketers can collect breadcrumbs to find my specific needs, preferences, and constraints, and can tailor messaging to me. As needs change, so does messaging to keep up with the changes. Once I purchase the smartphone, I should stop receiving those messages. Various touchpoints keep in touch with each other, so the call center and website are aware of advertising placed / responded to, and can customize their interactions based on the customer actions already known to them.

 
Source
< Prev   CONTENTS   Source   Next >