In this post, I’m exploring the usefulness of personas in the era of digital analytics. At Qatar Computing Research Institute, we have developed a system for automatic persona generation (APG) – see the demo. Under the leadership of Professor Jim Jansen, we’re constantly working to position this persona system at the intersection of customer profiles, personas, and online analytics.
Three Levels of Customer Data
To understand the benefits of personas versus other forms of online analytics, we note that there are three levels of customer data:
- customer profiles (individual level data)
- customer statistics (aggregated level numbers: tables and charts)
- customer personas (aggregated level individual data)
Which one is the best? The answer: it depends.
Let’s investigate a few use cases for customer analytics data.
Personas for Online Advertising
In online advertising, usually, the more individual the data, the better. The worst case is seen as the mass advertising, where there is one message for all customers: this fails to capture the variation of preferences and tastes of the underlying customer base, and the advertising becomes inefficient and expensive. Market segmentation (“Women 25-34, Finland”) performs better because it is aligning the product features with customer features on average. Using market segmentation to find communalities within the persona allows marketers to create more tailored and effective messages, which results in less wasted online ad impressions. Yet, the best scenario is having individualized data on customers, as this data can be used for personalizing the message’s content and format to each customer and, thus, achieve the highest level of match between supply and demand. Winner: individual level data.
Personas for Product Development and Design
Similar to advertising, product development and design involve a great deal of experimentation. Professionals have certain conventions that are adopted industry-wide in the long run (e.g., Amazon.com adopts a design practice and small e-commerce sites follow suit). In the product development game, many players prefer being followers, and this strategy works for the most part. However, it is crucial that the product changes are tested, as best practices do not necessarily generalize. Personas can be applied to challenge and evaluate items in the product development roadmap — e.g., “What would Joni think of this change?”.
In turn, multivariate testing using individualized data can reveal optimal designs better than personas that are “imaginary people“. Of course, personas, just like other immersive user-centric design techniques, can be used as a source of inspiration and ideas for product development and design. But personas are just one technique, not the only technique. Often, I would recommend experimentation with individual users over personas. A classic example of successful experimentation is Instagram that found from data that its filters were a killer feature. For such applications, it makes sense to define an experimental product feature set, and collect behavioral feedback from the users. Even though personas can also be based on user behaviors, the relationship between personas and predicting human behavior is weaker than collecting and analyzing real experimental data.
Coincidentally, business professionals often ignore systematic testing because they have a pre-defined idea of the user (cf. persona) and are not ready to see their ideas challenged (this is called validation bias). The more work is carried out to satisfy the imaginary user, the harder it becomes to opt for out-of-the-box design choices. Yet, those kind of radical changes are required to improve a product not by small margins but by orders of magnitude. Eric Ries call this ‘sunk code fallacy’. Therefore: Winner: individual level data.
Personas for Strategic Planning
For personas, I would say strategic decision making is very promising. Tactical and operational tasks are often better achieved by using either completely individual or completely aggregated data. But individual data is practically useless at strategic decision making. Here, aggregated data is a necessity (consider, e.g., sales by region or customer segment), and it is hard to see anything replace that.
In particular, data-driven personas can provide behaviorally accurate insights on the needs and wants of the market, and act as anchor points for strategic decision making.
Strategic decision aid is also a lucrative space for persona development; companies care less about the cost than in operational matters, because strategic decisions are of high importance. To correctly steer the ship, executives need need accurate information about customer preferences and have clear anchor points to align their strategic decision with (see HubSpot’s persona case study).
In addition, aggregated analytics systems have one key weakness compared to data-driven personas. They cannot describe the users very well. Numbers do not include information such as psychographics or needs, because such qualitative information is not available in numerical data. Customer profiles are a different thing — in CRM systems, attribute enrichment might be possible but again the number of individual profiles quickly becomes cognitively overwhelming for detailed analysis.
Conclusion of Personas’ Usefulness
The faster the movements towards real-time optimization, the less useful a priori conceptualizations, like target groups and personas, become for operational advertising, product development, and design. However, personas remain useful for strategic decision making and as “aggregated people analytics” that combine the coverage of numbers and the qualitative insights of individual customer profiles. The million-dollar question is: Is it possible to build personas that include the information of customer profiles, while retaining the efficiency of large numbers? At the Qatar Computing Research Institute, our persona team is working hard towards that goal.
See related persona research.
Jansen, B. J., Salminen, J., and Jung, S.G. (2020) Data-Driven Personas for Enhanced User Understanding: Combining Empathy with Rationality for Better Insights to Analytics. Data and Information Management. 4(1), 1-17. https://content.sciendo.com/view/journals/dim/4/1/article-p1.xml