Like the five blind men who cannot agree on what an elephant is, because each of them explored a different body part of it, market research sometimes has the same problem. It can describe parts of reality very accurately but is sometimes unable to see the big picture. In order to understand the whole truth, we need to get a holistic perspective on reality. In this post, we want to show how to get an integrated perspective on usage data.
Two types of data: a methodological challenge
Usually, market research deals with two different types of data. On one hand, this is data from reactive research, where the researcher interacts with the participant (i.e. surveys, focus groups). On the other hand, it is non-reactive research, where the researcher does not interact with the participant (i.e. observations). The problem is, that these two methodologies are very often disconnected.
Reactive research methods are perfectly apt for user data and usually describe the identity of people, e.g. demographic and social background, attitudes or opinions. It is relatively easy to carry out reactive research projects but you can only assess the research topics retrospectively. This is one of the many reasons, why reactive research is prone to biases: respondents have to understand the question, remember relevant information, find an answer to the question and finally express it in a comprehensible way. Reactive research methods work particularly well for user data.
Non-reactive research methods are apt for usage data and describe what people do, e.g. observable behaviour, activities or transactions. You always have to capture the data in the moment of truth, meaning you get authentic and potentially unbiased information in return. However, sometimes it’s hard to collect the data as you have to cope with a lot of complex information in real-time. Furthermore, interpreting the information might be challenging, as observational data does not tell you much about the meaning of the observed behaviour. All in all, non-reactive research methods are especially apt for usage data.
But why should we try to integrate the two data types? Above all, the difference between usage data and user data is merely analytical. In real life, there is no usage data without a user and vice versa . We just got used to the distinction of user and usage data, because our methods work accordingly. We either collect user data with our reactive methods or we capture usage data with our non-reactive methods. But reality is more complex: If we really want to understand the consumer, we need to bring both views together. An integral perspective can help to explain if and to what degree attitudes and previous experiences have an impact on consumers’ behaviours. The other way around it also helps to explain the impact of behaviours and usage on strengthening attitudes and creating experiences (e.g. brand loyalty, readiness to recommend a product).
Example 1: Product Tests
Let’s have an example. An easy way to explore the interplay between attitudes and usage is an In Home Usage Test (or Product Pilot as we call it) that consists of three stages.
- With our online panels, we screen users in the desired target group and ensure their availability for the project. At this stage, we collect user data by gathering feedback about their demographic background, previous experiences, attitudes and future expectations.
- We then start exploring the usage data. We pack and ship the desired product to participants and brief them how to document the usage. Depending on the product, this could be an online diary or a market research online community (MROC). At this stage we usually obtain a lot of pictures, videos and verbatim data.
- After the product has been used, we evaluate the usage experience with a survey. Among other aspects, the study could focus on user data again, i.e. if attitudes towards the product (i.e. likability to purchase) have changed.
Among many other things, the product pilot will allow you to explore the impact of product usage on attitudes and preferences. And it will allow you to identify the potentials for improving your product accordingly.
Example 2: Audience Analytics
In a digital world, we often have to evaluate the use of digital products, especially the usage of websites and interaction with digital advertising. In a lot of these cases, we use our Audience Analytics tool to capture the digital usage behaviour.
Audience Analytics is based on cookie tracking. Without going too much into the technical details, it allows us to identify our online panellists on every website that has embedded our tracking pixel. We will not only measure, when (time, date) and how (mobile, desktop browser) the visitor approached a specific site, but also know exactly who they are (user data). Thanks to the wide range of panel profile variables, we can assess the visitors of a website in real-time.
But that is only part of the story. The tracking pixel can be personalised with individual tags by the client to mark the context in which it is embedded. If these tags reflect the usage of the site (let’s say the amount of products in the shopping cart), you are able to trace back the user behaviour on a specific website (usage data).
It is important to mention that all data must be reported in an anonymised way to protect the privacy of participants in research. Very often the best way to do this is an interactive dashboard.
The importance of trust
Product Pilot and Audience Analytics are just two study types that incorporate user data with usage data. There are many more possibilities. But, there are also numerous challenges and pitfalls that need to be considered when integrating usage data and user data.
First of all, you need to get the consent of participants to integrate different data sources. They will only give their okay, if they fully understand the purpose of this data integration and, even more important, have a minimum level of trust. The point here is that you cannot really buy trust, it has to be earned over a long period of time. This is the reason why many companies fail at getting the necessary consent by their users. Fortunately, at Norstat we have long-lasting and trustful relationships with our respondents. It allows us to carry out projects, that otherwise would have been impossible, but of course it is limited to activities that don’t undermine the trust we’ve earned over the years.
Dealing with different data formats
Another challenge for researchers are data formats. Despite the fact that it is highly authentic, usage data is very often unstructured (e.g. images, videos, verbatim), extensive with regard to the data volume (e.g. timestamps, contextual data) and incomplete on a respondent level (i.e. measurement failure). Contrarily, user data usually is structured, standardized and complete for each respondent, but lacking a lot of topics that haven’t been actively covered by a questionnaire. You don’t get, what you don’t ask! Integrating these data formats, let alone the analysis of such hybrid data sets can be challenging. This is why we recommend using established study formats (like our Product Pilot or Audience Analytics), as they already cope with most of the problems mentioned above. In addition, online dashboards might be a simple and playful way to explore the complex data in an effortless way.
Benefits for brands
In our experience all these challenges are manageable. Researchers will start seeing the big picture once they take a holistic view on consumers and integrate user and usage data. It will help them to understand what customers think and how this relates to their behaviour. That in itself is already a huge benefit for researchers. But it goes even further: enhancing usage data from the client with user data from a panel will make it easier for brands to act upon the results since your data is not coming from the outside, but actively including the internal data sources of your client.
We should embrace the opportunities of integrating usage data with user data, as it will improve the understanding of consumers as well as the capability of brands to act upon research results.
If you want to learn more, feel free to get in touch with us.