Better customer insight – in real time

Better customer insight – in real time

Few marketing challenges are tougher than identifying and influencing what drives customers’ attitudes and behaviour. Traditionally, executives have relied on a combination of quantitative data from surveys and qualitative insights from focus groups and interviews.

Unfortunately, both kinds of research suffer from a fundamental flaw: they rely on customers’ memories, which decay rapidly and are often biased by context. Internet-based research tools suffer less from these problems because they can capture customer experiences almost immediately, before memory fades or becomes biased, but they can be used only with online interactions, which account for just 15% of customers’ encounters with companies and their brands.

The only traditional technique that really allows companies to record the complete range of customer experiences is ethnographic research, in which researchers shadow individual consumers and watch their behaviour. However, this is labour-intensive, expensive and potentially misleading: it’s hard to untangle the individual customer’s quirks from general customer behaviour. It also introduces another bias: the customer will probably have an unconscious desire to please the researcher. Corporations must therefore either rely on imperfect and biased memories or spend a lot to directly observe potentially unrepresentative behaviour. Either way, the insights and data on which they base their marketing decisions are inherently faulty.

Marketers have long sought a research method that can capture customer reactions immediately, does not intrude into those reactions, minimizes bias and can affordably be applied to customers in relatively large numbers. We believe that real-time experience tracking, or RET, a new research tool, rises to this challenge.

Designing the program

RET was born of two insights. First, while a market researcher can’t easily follow customers around 24 hours a day, those customers’ mobile phones can, and unlike human observers, they don’t sway people’s perceptions of experiences. Second, although customers may interact with a company in thousands of ways, you really need to know only four things about each encounter: the brand involved, the type of touchpoint, how the participant felt about the experience and how persuasive it was.

We developed a quick microsurvey that customers can take on their mobile phones every time they encounter a company’s brand. The survey requires participants to input a four-character text message.

Participants completed four phases of research:

1. They filled out an online questionnaire about their awareness, knowledge, perception and use of the company’s brand or product and those of four or five competitors (without knowing which firm commissioned the research).

2. They texted a four-character message whenever they came across any of the brands over the course of the research project.

3. They were asked but not obliged to keep an online diary in which they expanded on their encounters with the brands and how they felt about them.

4. At the close of the project, they completed a modified version of the first questionnaire to see whether their attitudes toward the brands in question had shifted.

Challenges and limitations

To ensure a balanced, representative sample, you need to profile potential respondents through a series of questions. A sample that matches the firm’s target market on demographics and other relevant criteria can then be assembled. Other factors to consider are the respondents’ marketing literacy, assertiveness, shopping enjoyment and confidence in obtaining information from peers online or offline. If a sample is still skewed following those checks, the results can be adjusted by giving greater weight to responses from underrepresented sections of the target market, just as in political polling. In highly diverse sectors, a larger sample is sometimes needed so that different segments can be analysed separately.

In setting up the text surveys, it’s important to provide a reasonably comprehensive list of touchpoint types, covering both direct encounters and indirect ones.

Participation in the surveys inevitably raises respondents’ awareness of the product category during the study period. So the purpose of the second questionnaire is to unearth the relative changes in respondents’ awareness, knowledge, perception and use of the various brands or products. You can also deal with the problem by taking a randomized control group from the initial sample. The participants in this group skip the text messages, simply filling in an adapted survey at the start and end of the study period. Shifts in their attitudes or key behaviours over that time frame can then be compared with those of the main group.

What the data tell you

A benefit of tracking an individual over time is that you can often cover a complete customer journey from the identification of a need to a purchase. And with statistical tools, you can use RET data to identify not only what most motivates customers to buy your brand but also how various touchpoints combine in a chain to influence the customers’ decisions. Here are some useful analyses you can conduct:

Key drivers. Applying simple regression analysis to the RET data can quickly tell you which touchpoints are most closely correlated with individual customer behaviours, such as a request for more information or an actual purchase. A good way to present this information is in a diagram called an odds analysis, which quantifies – and compares – the relative likelihood that various touchpoints will lead to the behavior in question.

Competitive analysis. You can also see how effective your touchpoints are at driving behaviour and shaping attitudes relative to the touchpoints of your competitors. A good way to present this is with a touchpoint impact matrix, which compares the performance of multiple touchpoints.

Chains of touchpoints. Customers’ reactions to each exchange are shaped by their previous interactions with a brand. The decision to go into a store might result from a conversation with a friend, seeing the store window display or something else. RET data can indicate where such chains might be broken.

Insight to action

Insights gained from RET can be acted on immediately – a great advantage in new product launches or marketing campaigns conducted in fast-changing environments.

Responding to RET findings in real time is easier in theory than in practice, however. That’s because RET covers the complete customer journey, which means the data it generates are useful to virtually every customer-facing part of a firm – from marketing communications and PR to operations and service delivery. Reaching, mobilising and co-ordinating all the relevant decision-makers therefore presents a huge challenge, especially for products and services that are promoted and delivered through multiple channels over large areas.

As market researchers, we’ve found an emphasis on behaviour. Real-time experience tracking, which goes beyond recording behaviour to help us gain insight into the rich perceptual and emotional worlds human beings live in, helps redress the imbalance. And because RET enables companies to assess and respond in real time to customers’ reactions to products, services, or branding efforts, it can play a central role in allowing customers to help design their own experiences with products. As RET and tools like it emerge, we expect that marketing will cease to be a game of stimulus-response and will evolve into a continual process of co-creation.

Emma K Macdonald is a senior lecturer at Cranfield University’s School of Management in the UK. Hugh N Wilson is a professor at Cranfield’s School of Management. Umut Konus is an assistant professor at the Eindhoven University of Technology in the Netherlands.


Notify of
Inline Feedbacks
View all comments