#InnovationSnack – How to profile a customer from A to Z
“As soon as I enter my favorite coffee bar, the barista already knows the coffee I want – not too long and not too hot: he brings it to me without asking, while we have a small talk. And I prefer going there rather than the coffee machine, and I would do it even if his coffee was less good or more expensive than the machine one. This is nothing more than a form of behavioural profiling: in short, nothing new beneath the sun”. Serena Torielli, CEO, and co-founder of Virtual B, opened the webinar held on 7 May on the Zoom platform with a provocative statement, in which she spoke about Behavioural Profiling1 together with Professor Flaminio Squazzoni of the Behave Lab at the Università Statale di Milano and Raffaele Zenti, Head of Data Science and co-founder of Virtual B.
In fact, as Squazzoni confirms, ‘the ability to “profile” those in front of us, in order to send clear and targeted messages, has always existed, and is one of the characteristics that have enabled men to distinguish themselves from other species. And it is still widely used today, at all levels: for example, it is not only the company that profiles the client, but also the opposite.
So why are we talking about it?
What is relevant here is that we need to teach this “behavioural profiling” to a machine, starting with a huge amount of raw and inhomogeneous data, while applying it to the financial sector, traditionally far from this technological and psychological use.
Although at first glance, what a barista does by remembering his customers’ tastes and preferences may seem trivial, the work that occurs inside his head is incredible – the outcome of more than a thousand years of evolution. And transferring these skills to a machine, in order to translate it into action on various service channels (financial and insurance advisors for example), is not a simple task.
How is this done?
In order to profile an individual, in fact, we need different types of information, all of them are essential: beliefs expressed, data relating to concrete choices, information about the social dimension and so on. For example, Squazzoni observes, “measuring human behaviour only from the digital traces point of view left by an individual during his online navigation would be misleading”. This is because actual choices are always “constrained” by other factors, which can be socio-cultural conditioning, economic necessity and so on and so forth. By integrating this data with spontaneously provided answers, for example through a survey, it is possible to grasp what the individual – client or prospect – would have done in optimal and unconstrained conditions – in other words, we get to know their true desires.
At the same time, as Zenti agrees, “you have to take into account the fact that when a person answers to a survey they may do it quickly, with their head on the clouds, or perhaps without being completely honest. So, you also have to look at direct behaviour: customers go to the bank or insurance website, open or don’t open the newsletter, sell or don’t sell financial instruments in some market situations”, all actions that tell us something about them.
Then, this data must be “cleaned” and, subsequently, integrated with broader data – for example, from public databases or surveys, which allow us to “pigeonhole” the individual into what in marketing are called the “persona”: “in Virtual B, we have over 50 models that include factors, such as propensity to feel part of “a tribe”, propensity to use digital tools and much more”, recalls Zenti. At this point, there is a long process of integrating heterogeneous data to obtain a coherent picture that can be used by an algorithm.
Finally, there are precisely algorithms, “which we put together by trying to use the right pliers'”, Zenti concludes. “The result is to mimic the barista that offers the coffee, but getting a machine to do it is very complex”.
From connections to relationships
This is, in a nutshell, behavioural profiling: it is an important technique, because allows financial companies to have a 360-degree view of their customers and it offers the possibility to personalise services and products and to communicate the methods, style and timing appropriate to each customer.2
“And then it is important for the bond creation – because a connection is a very different thing than a relationship,” adds Serena Torielli. “A platform such as ours, can turn a connection into a relationship.”
Because eventually- quoting Ennio Doris – what people want, and ultimately deserve, is just a little bit of attention.
The Virtual B solution
Virtual B has been working for years in the financial sector, in close contact with data and its analysis. Our experience has been creating different solutions that generate value and solve problems for financial and insurance intermediaries.
If you are curious about this topic, contact us at the link below for an explanatory demo and discover how to apply our Behavioural Profiling logic to your business processes.
1 – Behavioural Profiling: going beyond traditional approaches
2 – Behavioural Profiling: an emotional intelligence tool for modern advisors