Everything begins with semantics
Before we dive into technicalities, let’s align on the wording. Segments, profiles, archetypes, and personas are frequently used, abused and confused. From our perspective, they all provide an insightful view of the targeted user and context. However, we spot some conflicting demands when looking at how these seemingly similar terms are used in different disciplines.
Be warned, the following is probably enough for hefty semantic discussion, but hear us out:
When using user profiles, marketers are looking for a face to represent their quantitative demographic market segmentation, aiming to better target and personalise. Their needs contrast ours as designers, as we look for a more effective user-centric way to innovate with surprising solutions.
It is time to clearly distinguish “market segments” from “need-based personas” to avoid future confusion. While the ‘upper-class urban professional’ is a term we often see being used for personas, it is understood already as a market segment, given the demographic nature. On the other hand, our need-based personas often have names like ‘the supporting caretaker’, as that frame provides much more space for deeper insights on behaviour and drivers.
At Koos, we use need-based personas in our design processes
How did we come to this research method? When defining need-based profiles at Koos, we use morphological psychology to uncover (sub)conscious needs. By basing our personas on this framework, we develop them as next-level empathy builders in an organisation and a powerful tool in the design process. These need-based personas came alive through explorative research, such as in-depth interviews and generative techniques. This way, we can provide insights into your users’ needs, motivations and expectations, which are impossible to extract from quantitative data (such as demographic or survey data) alone.
Imagine the scenario of developing a new digital pharmacy service …
A client approached us with a twofold challenge: discovering patients’ needs for ordering and receiving their medicines as well as identifying the patient group with the highest potential to use the new proposed online service. Creating need-based profiles was a great way to get an understanding of the context and users’ needs. However, after understanding the scale of needs of the potential customers, questions arose regarding the size of certain patient groups and how to target them: To what extent are these patients, frequent medicine users? How far do they live from a physical pharmacy? Questions we could not answer with qualitative research only.
Combining the best of both worlds: data analytics and service design
Seeing the power of combining data with service design pushed us to a new perspective on our processes. It turned out we needed data-infused profiles. Data-infused profiles are composed of quantitative as well as qualitative insights, connecting both the behavioural and demographic data to the needs, emotions, and motivations of the user.
Most significant benefits of data-infused profiles
- Identifying surprising patterns in quantitative data that one cannot detect with the unaided eye.
- Understanding the proportion of needs and each profile in the total user population provides greater certainty in (strategic) decision-making.
- Empathising with users by understanding their latent needs, emotions and motivations.
Referring to the previous online pharmacy example: on top of knowing the users’ key needs, additional (demographic, behavioural) information can support targeting the user. For example, when you discover that people interested in an online pharmacy service often live relatively far from a physical pharmacy, you can consider developing new services for these specific people. So, besides having insights into user needs, we can now also target design solutions to the right person.
Creating data-infused profiles: rule-based versus cluster analysis
Let’s break this down further. We identified two approaches to define user profiles based on data: 1) rule-based or 2) clustering techniques (using Machine Learning). Rule-based methods can validate predefined (insights and assumptions of) user profiles by grouping data based on predefined thresholds or rules, resulting in predictable outcomes.
Applied to the online pharmacy project, we predefined two groups of patients: potential users and people who are not interested. A potential user is one that 1) is often ordering online and 2) has an intrinsic interest in ordering medication online. Based on patient questionnaire data, we defined clear rules to divide respondents into these two profiles.
An alternative approach is to let the data speak for itself. However, as humans do not speak data, we must rely on the almighty computer. We use clustering techniques and apply mathematical models to discover similar individuals in a dataset and group them into different ‘clusters’. It is an unsupervised method because predefined rules or characteristics do not guide it. In this case, data was used to explore and generate the foundation of the profiles.
As an example, we performed a cluster analysis for a global leader in the HR services industry. From prior research, we were able to define 10 key needs. Based on a questionnaire assessing these 10 different customer needs, the clustering algorithm suggested 3 different profiles. However, what do these differences actually mean? To make the profiles usable, we still needed to interpret them. For that, we again relied on insights from in-depth qualitative interviews.
Which method is most appropriate, you ask? As always, it depends on the available data, the scope of the project, and the specific purpose of the profiles.
Make better decisions with the user at heart
Genuine empathy is crucial for success
Having access to lots of data is great, but it is not all. As Jeff Bezos said: “When the anecdotes and the data disagree, the anecdotes are usually right. There’s something wrong with the way you are measuring it.”
Data is not a replacement for talking to real users. Actual insights come from seeing the world through someone else’s eyes. Therefore, we always ensure that real users are closely intertwined with our design process, even when leveraging heaps of quantitative data. Bezos, again: “You collect as much data as you can. You immerse yourself in that data… but then make the decision with your heart.”
Working data-driven with the user at the heart
Data-infused profiles can support organisations in making better decisions with the user at the centre, whether it is about directing innovation efforts, product portfolio, or strategic positioning. Mixing qualitative and quantitative methods can create reliable and inspiring profiles simultaneously. Ultimately, we believe that well-created data-infused profiles can act as a springboard for incremental as well as radical innovation.
At Koos, we continuously develop our tools and methods to serve our clients’ needs best. Want to know more? Get in touch!
Reach one of the co-authors if you want to know more about data-infused profiles.