Customer analytics and market research – the two areas that strives for better understanding of customer needs and preferences. Nevertheless, they are organised in separate departments and attract very different people. Why they operate separately? And how the organisations ensure that they fully benefit from all sources of customer insight?
Working in data analytics capacity within the market research agency along with quant and qual market researchers has been certainly an inspiration for me. Well, I still believe that data on customer transactions and behaviours should be at the centre of all marketing activities. But this work experience has made it more obvious to me that both customer analytics (sitting often within CRM team) and market research (managed by Insights or Marketing teams) can benefit from working closer together.
The market research as a way of getting an understanding of consumers' wants, needs and beliefs has had around a century to establish itself and earn its status among marketing and insights managers. But this is not the only reason why the market research agencies became an extended hand of client side marketing departments. Marketers feel more confident in this area. Market researchers simply speak their language. Market research is all about people and customers starting with focus groups, interviews through to online forums, customer panels and respondent surveys.
Customer data analytics has developed on the back of technological progress in data collection, cheaper data storage and growth of accessible computational resources. It has been getting more attention in the recent years as part of ‘big data’ buzz. Companies are upgrading their CRM systems, building data warehouses and establishing analytical capabilities. Few have made significant progress; many are still at the stage of reporting rather than extracting insights from the customer data. Strategic use of customer data poses for front-runner marketers an exciting opportunity, while for others it is an inconvenient step into the uncharted territory.
Quant market research and customer analytics are the two major sources of customer insight. Nevertheless they tend to be organised in different teams or departments. The way they communicate or work together (and with marketing) vary across companies. On the agency side, there are not many that offer both market research and customer data analytics, though some market research agencies are seamlessly adding data intelligence capabilities to their offer.
Customer analytics has emerged as a separate source of customer insight parallel to market research rather than branching from it. Both, customer analytics and market research use analytical techniques to extract insights from data. These insights provide valuable input for developing brand strategy and can be instrumental for its execution. Given that the ultimate goals are similar, why are these two areas organised separately?
What is it that sets data analytics apart?
What Makes Customer Data Analytics Different from Market Research?
There is actually a number of distinctions. The list below is not exhaustive and aims at capturing only major differences.
Survey vs Actual Transactional Data
The most obvious difference lies with the input data used. While market researchers rely on data collected primarily through surveys, data scientists have to find the ways how to answer questions using available transactional and other behavioural data. Alternatively, (s)he needs to convince stakeholders that more detailed data are essential to answer respective business questions.
This has several implications
Growth of Transactional Data
First of all, the origin of data – survey vs customer and transactional databases - has implications for the size of datasets
Survey data are most likely to stay contained. The budget is here the main factor naturally constraining research data proliferation. Even if some companies made progress in building own customer panels, there are limits to how much useful attitudinal data they will be able to collect.
On the other hand, the size of behavioural and transactional databases are growing and becoming more granular. Steadily declining storage and computational costs are significantly contributing to this trend.
Data- vs Research Design-Driven Approach to Problem-Solving
Market researchers have more freedom in their research approach as they determine what data and how they will be collected through respondent surveys. Full ownership of research design allows more sophisticated methodologies developed for tailored datasets.
The data in customer databases can be abundant but quite basic. Therefore, even simpler manipulations of large data are often sufficient to provide wealth of insight. As this side of business is becoming more comfortable with data processing, they are likely to go into building richer datasets and more accurate predictive models.
Market research takes pride in attracting people with diverse backgrounds, including market research, marketing, psychology or sociology, anthropology, just to mention a few. There is very strong focus on communication skills. Analytics professionals tend to have more quantitative and technical background – econometrics, statistics, operations research, computer science or data science in general. Rigorous theoretical and practical experience is a must to understand stats and numbers. As a result, the cultural difference between people in market research and marketing analytics can be quite significant.
Project Result Implementation
In analytical projects, similarly to the market research, the output can be communicated through a set of sleek slides delivering the main results and recommendations. However, the outcome of analytical project is often a predictive model, i.e. a set of rules or an algorithm that enables projecting future developments or a propensity of customers to certain type of behaviour in the future. Models are built with a specific objective in mind. Model implementation is therefore an important part of the process. Projecting the most likely next step of individual customers helps the business prioritize resources and taking the right steps at the right time. Market research that has broader strategic focus should provide guidance on what sorts of behaviours the built models should be predicting.
Analytics as Part of Continuous Process
Closely related to the point above, analytical project cycle is continuous. The built model is implemented to provide input for decision-making; the results are tracked and validated. The model is updated in regular intervals or when accuracy of predictions declines. Some models become irrelevant when the overarching business strategy changes and are replaced by new ones. Again, this is different from market research that helps answering those questions for which behavioural and transactional data are not available.
Embrace the Difference
The list of contrasting aspects of the two areas and those working in them: data scientists and market researchers, may build a case why they should operate separately. But this is a hasty conclusion. Given the common goals, all areas of customer insights should embrace diversity and learn how to work together.
Market research is perhaps more prominent (at least so far) when it comes to designing and testing new product concepts, value propositions or studying underlying needs and motivations that drive customer preferences. Nevertheless, data should be at the core of brand strategy when it comes to customer experience, new customer acquisition and retention strategies, media optimisation, customer journey improvements and personalised targeting, to mention a few.
The two areas provide different angles of looking at customers. And as such, an overlay of the two is required to get below the surface. The outcome of attitudinal research requires the context of transactional and behavioural data to put research findings into action.
For example, a research study finds a gap in customer needs and develops a new service concept. Customer data are important to identify the customers who would be the best target for the new service.
Similarly, behavioural analysis can be leveraged by attitudinal and need research getting more insights into the underlying reasons behind certain behaviours. So, for example if a customer data analysis identifies bottlenecks on the customer online journey, survey-based research brings more “Why”- type of insights. With this approach, brands can benefit from customer-driven innovation.
The quality of input for strategic decisions depends on how closely different areas of business intelligence co-operate. The siloed approach is by no means sustainable. Even if individual teams deliver excellent jobs, they cannot perform to their potential while working in isolation. Working as one team or fostering co-operation requires strong communicators who see the benefits of synthesizing the output from all insight sources. And, of course, who enjoy talking their walk.
10 April 2015, Sydney,
Elena Yusupova, PhD