Kihyun Hannah Kim
Indian School of Business
Direct marketing communication is often seen as unnecessary. Nonetheless, Hannah Kim and V Kumar ask: Can it still help firms to make money? How can they communicate with each customer to increase its effectiveness? Is it possible to determine whether money incentives or relationship building is more effective? More importantly, can firms discover their customers’ preferences without asking?
Due to the proliferation of marketing channels in the digital world, the decisions of C-suite level executives have become even more complex. “The concept of finding the right customer at the right time with the right message still exists,” says Kevin Akeroyd, CEO of Cision, a platform provider for PR and earned media. Yet many companies struggle to find the most effective direct marketing strategy to convert and retain customers. Furthermore, proving the return of investment (ROI) of marketing initiated by the firm is still, for most, a critical challenge.
We gathered the following examples during interviews with executives. These two messages were sent from business to business (B2B) sellers to their customers:
The next two messages were sent by a telecommunication firm in the business to consumer (B2C) context:
Far too often, managers need to answer a variety of questions before implementing a direct marketing strategy. When should each type of message be used? Should the monetary message (i.e., 10 percent discount or $100 rebate) be sent earlier or later in the relationship? And the relational message (i.e., help grow your business or understand your usage)? To which customers should it be sent? When should it be sent? How many relational messages should be sent before switching to monetary messages?
Managers who make decisions about which message to send find themselves at a crossroads; should they offer monetary incentives or relational benefits?
Firms send direct marketing communications or targeted messages continuously in order to build strong relationships with their customers. It is only natural, then, to question whether we are spending the right amount of money on the right kind of direct marketing communications for each customer. Managers who make decisions about which message to send find themselves at a crossroads; should they offer monetary incentives or relational benefits? According to Gartner Research, most firms spend about $12 billion each year to manage customer relationships, minimizing customer churn and cultivating profitable interactions.1 B2B firms typically use emails, phone calls and in-person conversations to communicate the monetary and relational value they can offer to each customer. B2C service and product firms also spend a large part of their budget on marketing in order to retain their existing customers.2 However, for the managers at these firms there are still no clear answers about direct marketing strategy. Beyond the knowledge, drawn from experience, that it is important to deliver relational marketing communications early and monetary marketing communications later,3 it is difficult to find any established rules for dynamic marketing (budgeting). Previous academic research has largely overlooked the content of direct marketing communication and focused only on channel and total marketing investment. Changes over time in the effectiveness of each communication to each customer have not been studied in depth. Further, more than three-quarters of global businesses still report that they do not take advantage of technology to manage customer relationships or to better understand which marketing efforts work.4
Because each customer’s experience with a firm is different, not all customers require the same level or type of marketing communication. To build strong and profitable relationships, firms need to tailor direct marketing to suit their customers’ preferences, help foster positive perceptions about the firm, and influence purchase behavior, all of which can help to improve the firm’s financial performance. One of the most common methods of understanding what customers want is to ask them. This approach may not always succeed, however, since what customers say they care about may not always be what they actually want.5 How, then, can marketers determine the most appropriate type of marketing messages without asking?
Our research reveals that, by classifying the core value emphasized in each past marketing communication as well as how customers responded (as evidenced by transaction data), firms can determine the ‘right’ type of message for each customer in future.6 We found that by tracking how customer preferences change over time, marketers can customize their communications to each customer and improve their firm’s financial performance.
Our research answers the following questions:
- How does the influence of monetary and relational marketing communication on customer purchase behavior vary over time and across customers?
- How can firms understand their customers’ direct marketing communication preferences without directly asking each customer?
- How can firms use evolving information about the type of message each customer prefers to improve their overall financial performance?
To answer these questions, we tracked four firms’ marketing input and transaction data at each customer level. We included a high-tech firm and a logistics firm in the B2B sector as well as a retail bank and a telecommunication company in the B2C sector. The sample of customers per firm ranged from 550 to 780, while the data were available from between 3.5 and 4.8 years. To provide more detail, we use the example of a Fortune 500 B2B service firm where we used data from 675 customers over a period of four years. We can categorize direct marketing communications as either monetary or relational by using text mining to analyze them empirically. We then examine the relative influence of monetary versus relational direct marketing on each customer’s purchase decisions, revealing which type of communication each customer prefers. Using this information, we provide guidance to managers on how to consider their customers’ evolving preferences in order to best allocate a given marketing budget.
Looking only at the total marketing investments over time is not sufficient to understand the customer’s purchase decisions because different types of messages have different effectiveness.
Understanding What Customers Want
Figure 1 presents the behavior of two customers (A and B) over the four year observation period. Both have begun their relationships with the focus firm at the same time and represent firms of similar size, annual revenue, and expected purchase volume.
However, as shown in the figure, the monthly sales revenue for these customers differs noticeably over the four year window. While the focus firm saw a significant decrease in revenue from Customer A, its revenue from Customer B grew considerably. Although the firm’s investment in marketing to each customer was similar, the proportion of monetary versus relational marketing differed significantly.
Customer A – Direct Marketing Communication & Sales Revenue
Customer B – Direct Marketing Communication & Sales Revenue
For example, monetary marketing communication to Customer A was curtailed in the later periods, which coincides with the decrease in her purchases. Meanwhile, the increase in Customer B’s purchase expenditure can be linked to a steady increase in relational marketing communication from the middle of the observation period on. This correlation suggests that direct marketing communication has a long-term effect on the customer’s purchasing behavior. Looking only at the total marketing investments over time is therefore not sufficient to understand the customer’s purchase decisions because different types of messages have different effects. Furthermore, since the relative effectiveness of monetary and relational marketing to each customer can vary over time, this example also demonstrates the need for a more sophisticated method of understanding what customers actually want.
To understand how the firm’s efforts result in different responses from customers, we first examined what communications the B2B seller sent to each customer. We define monetary marketing communications as including entirely economic aspects of direct marketing (e.g., promotion), which are then evaluated by the customer’s rational judgment. This marketing focuses on making the relationship more financially attractive so as to create financial bonds with customers. Relational marketing communications, by contrast, are messages which focus on non-monetary worth (e.g., support services), and which evoke emotional responses. These efforts strive to make the relationship more personal and socially attractive and to create social bonds. We thus examine what has been communicated to customers by a qualitative analysis of the marketing communications written by the B2B seller’s employees. Using text-mining, we identify the core value offered in each contact and then measure empirically how much the firm invested in that particular type of value for each customer, each month.
To discover how these direct marketing communications influence customers’ purchase decisions in the long term, we used a popular statistical technique called state-space modeling. This approach allows us to capture the evolution of ‘state,’ the direct marketing preferences of each customer, over time. In simple terms, customers’ responses to current marketing may be influenced by messages that they have received in the past. We therefore include past experiences in the model so we can understand how firms should implement different marketing strategies at the right time.
By using a state-space model in a Bayesian framework, we were able to trace each customer’s sensitivity to monetary and relational marketing communications.
Figure 2 shows an overview of the modeling process.
We compare the relationship between the monetary and relational marketing communications with sales revenue and find that customers’ preferences evolve as they accumulate experience with marketing communications.
By using a state-space model in a Bayesian framework, we were able to trace each customer’s sensitivity to monetary and relational marketing communications. We corrected for the potential endogeneity bias, that is the possibility that monetary and relational marketing may be applied in response to the customer’s previous purchase level, by using the firm’s past marketing budget strategies as an additional instrument.
We further validate our predictions through a comparison with what customers report about their preferences, thereby determining the extent to which our inferences align with what customers say. Since customer surveys are conducted very rarely (e.g., annually), they fail to capture the continuous changes in customer preference. Nevertheless, self-reporting can be useful in determining customers’ overall satisfaction, information which we use to validate our model predictions. Having discovered the ‘state’ through this modeling process, we can then compare the relative influence of monetary and relational marketing on purchase behavior. This comparison yields a clear picture of the type of direct marketing communication that each customer prefers at any given time, their heterogeneous preferences.
Next, we ran a simulation analysis to determine how much revenue the firm could generate by allocating marketing resources based on the changes we predicted in customers’ preferences. By reallocating the marketing resources devoted to each customer from our sample in accordance with the changing preferences we traced, we found that, with only 675 customers, the firm could increase its total sales revenue by 8.3 percent (about 2.5 million dollars) over the four year observation period.
Figure 3 shows how businesses can implement our approach.
Many firms could use this same procedure. After the B2B service firm, we tested our method of customizing direct marketing communications in three other firms including a B2B high-tech firm, a retail bank, and a telecommunication company. Although relational marketing by phone or in person is relatively rare in consumer markets, B2C firms face similar challenges when marketing through direct mail and email. For example, the managers at a retail bank wanted to decide which kind of message to send: “introducing a low cost service fee of 0.35 percent on your investments” (i.e., monetary marketing communication) or “we want to help you meet your future goal” (i.e., relational marketing communication). We used the steps in Figure 3 to calculate the future preferences of the retail bank’s customers as well as those of customers of the B2B high-tech firm and the telecommunication company. We found that, by sending out targeted marketing based on our predicted customer preferences, the sample firms would increase their average annual sales revenue per customer by 2.1 to 3.2 percent (between $2.2 million to $4.1 million total) over the four year period. With a 5 percent increase in spending on tailored relational or monetary messages, these firms’ sales revenue would increase by more than 20 percent.
Successfully Implementing the Direct Marketing Communication Strategy
By scrutinizing individual preferences for monetary or relational value, our study also identifies links between marketing preference and customer characteristics. Based on our findings, we have divided customers into four broad groups: the ‘show me the money’ group, which is most interested in offers of monetary value; the ‘let’s be friends’ group, which is enticed by offers of relational value; the ‘give me everything’ group, which wants to get all the value offered; and the ‘leave me alone’ segment, which is unlikely to be persuaded by any direct marketing communications (See Figure 4).
We use the size of the firm, its sales revenue, and information about its industry to understand B2B customer characteristics. For B2C firms, we use demographic information and customer transaction behavior to understand the associations between customers’ characteristics and their preferences.
We have discovered four key findings that have important managerial implications:
Small firms prefer direct marketing that emphasizes relational value.
Among B2B customers, we found that smaller firms (i.e., companies with few employees) generally want to build stronger ties with the focal firm and thus prefer an emphasis on relational value. Among the customers of one of the B2B firms (which deals with small and medium sized clients), we found that ‘let’s be friends’ firms employed an average of 8.4 people, whereas ‘show me the money’ firms averaged 16.0 employees. This result contradicts the common argument that small firms prefer to maintain their flexibility by remaining less involved in relational exchanges. On the other hand, compared to smaller firms, customers with high purchase expenditures (e.g., large firms) value both types of direct marketing communication more equally. Cognizant of the interplay between monetary and relational marketing, these customers prefer a mix of the two.
Customer demographics influence choices about relational vs. monetary value.
Among customers of the B2C firms, more high income (annual income > $110,000) customers belonged to the “give me everything” group. Older customers (age >50) were often in the “show me the money” group, while women were often in the “let’s be friends” group.
Preferences vary by industry.
We found that B2B customers in the automobile and health industries most often belonged to the ‘show me the money’ group, while those in the retail and service industries more often belonged to the ‘let’s be friends’ group. B2B customers in high-tech and manufacturing industries were most often in the ‘give me everything’ group. So by selectively targeting firms with particular characteristics, rather than allocating marketing investments to randomly selected customers, B2B firms can expect positive responses.
More is not necessarily better and timing matters.
We found that, for both B2B and B2C firms, the majority of the effects of direct marketing communication dissipated after a few months. This result is consistent with the notion that, while communicating value to customers helps to shape the relationship between the customer and the firm, repeated interaction is necessary to prevent the effects from diminishing. However, we also found that some customers prefer not to receive any direct marketing communication and may respond negatively (i.e., reduce their purchases). Some customers prefer to build strong relational ties in the beginning and to discuss monetary value later on, while other customers prefer the reverse. The timing of this shift varied significantly according to past marketing experiences. So when a firm is deciding what type of direct marketing communication to use, we found that it should first ascertain how much monetary and relational marketing has been sent in the past in order to determine what to emphasize next.
Managers can continually update their classification of customers by knowing their preference as those preferences shift over time.
As the interactions between customers and firms become ever more complex, the importance of gaining a competitive advantage by finding the right direct marketing strategy increases. By tracking the content of direct marketing communications according to the monetary or relational value offered, managers can improve their firm’s overall financial performance by strategically allocating marketing resources over time. In applying the right type and quantity of direct marketing communication, firms can expect to improve the responses from each customer and reduce unnecessary marketing investment. Furthermore, managers can continually update their classification of customers by preference as those preferences shift over time. This improved understanding of customer preferences can lead to greater efficiency and profits, and ultimately provide a competitive advantage. Because rich data has become more accessible, B2B practitioners have begun to look for guidance in using the available data to address business problems.7 Especially in light of the growing use of artificial intelligence (AI), such as machine-learning technologies and text analytics, firms can use the proposed framework to analyze their direct marketing messages, capture shifts in their customers’ preferences and immediately adjust their marketing strategies.
Kihyun Hannah Kim is an Assistant Professor of Marketing at the Rutgers Business School, Rutgers University. She earned her doctorate in marketing from Georgia State University. Her research focuses primarily on customer relationship management, marketing communication, social media, and online marketing strategy using empirical models and large-scale data.
V. Kumar (VK) is Distinguished Term Professor and Senior Fellow at the Indian School of Business, HUL Visiting Chair Professor, IIM Ahmedabad, India; Distinguished Fellow of MICA, and Distinguished Professor of Research at WE School. VK was Editor in Chief of the Journal of Marketing and named a Legend in Marketing. VK has published many books and hundreds of articles in journals including the Harvard Business Review, Sloan Management Review, and California Management Review.
1. Market Snapshot: Customer Relationship Management, Worldwide, 2012. 2013: Gartner Research
2. Highlights and Insights Report. 2017. Retrieved from https:// https:// cmosurvey.org/wp-content/uploads/ sites/15/2017/05/The_CMO_Survey- Highlights_and_Insights-Feb-2017.pdf.
3. Altman, I., Negotiation: What Makes the Right Business Deal, Forbes. 2015.
4. Duffy, S., High-Value Marketing: Connecting Customers through Technology, Analytics, and Collaboration. 2016: Harvard Business Review Analytic Services Report.
5. Boaz, N., J. Murnane, and K. Nuffer, The Basics of Business-to-Business Sales Success. 2010: McKinsey & Company.
6. Kim, K.H. and V. Kumar, The Relative Influence of Economic and Relational Direct Marketing Communications on Buying Behavior in Business-to-Business Markets. Journal of Marketing Research, 2018. 55(1): p. 48-68.
7. Lilien, G.L., The B2B Knowledge Gap. International Journal of Research in Marketing, 2016. 33(3): p. 543-556.