Data and digital innovation: how much are they shaping the financial industry?
It is widely believed that data-driven technologies and digital innovation can provide financial and insurance services at a cheaper price. Efficiency gains would be beneficial to the profits of an industry that is proverbially not very dynamic (just think of the various Paleozoic terms still used, such as Cobol1) and that is experiencing a period of great margin compression. There would also be many advantages in terms of financial inclusion and productivity gains in other fields of the economy.
But how big is the productivity gain related to digital innovation and data usage (i.e., Machine Learning, AI)? In short: with operational models achievable at a reasonable time, how much can we gain?
Let’s try to calculate the possible returns thinking of wealth management in a broader sense, that is investment services, protection of people and things (insurance) and loans. Namely, the current offer of banks and insurance groups.
Let’s just think of one of these situations, using the classic business model that is generally based on an established network of professionals – financial advisors, private bankers, insurance agents. Let’s just consider a technology that does not replace these Sapiens, but allows them to make better and faster decisions, by automating certain tasks. In practice, a technology that:
- precisely identifies customer needs and objectives directly from the data;
- helps creating the best tailor-made mix of products and Next Best Action;
- automatically offers customised content, forms and reporting;
- follows the customer along his lifecycle, starting from the engagement phase, but always learning and adapting to the different cases.
Now, let’s ask ourselves: what happens if we provide the company and its network with this technology (which, incidentally, is Virtual B‘s core business)?
To answer, we first gathered information from financial advisors about their operability and time saturation, the quantity and quality of customers, how they prepare to meet them and other process information. In short: times and methods. We have considered the data on the distribution of capital capacity among customers, integrating some of their databases. Since there is a variety of answers and little certainty in this world, a probabilistic model has emerged, calibrated on the information collected and the data available, which describes the customer management process. It is based on the Queueing Theory2, a field of applied mathematics that is quite popular within the analysis of telecommunications networks and which also lends itself to represent the financial/insurance consulting service. Simulating everything with the Monte Carlo method and synthesizing the shapeless cloud of tens of thousands of numbers, lots of interesting facts have emerged.
If you introduce this technology on a network whose production capacity is not far from its saturation point, the software allows a significant time saving of 35% on average (conservative estimate, just think of the time needed to assemble a decent report or study the customer profile). What about the productivity gain, understood as more manageable customers? The result is surprising: with a probability above 90%, the ability to manage customers doubles. Doubles, yes. And sometimes it triples.
It may be surprising that a technology that leads to a saving of about a third of the time leads to more than twice the number of manageable customers. Apart from the fact that few things travel in a straight line in nature and even less in economics, this is typical of systems presenting process inefficiencies and “bottlenecks”. Furthermore, data science and customer intelligence go hand in hand and, yes, small miracles happen.
However, new customers to serve are equivalent to new masses under management or new premium income. But, since customers obey to a Pareto law like the one in the following chart (estimated on the real population), where very few have much and many have little, this probably means that you are going to deal with customers with less potential. Usually consultants and agents focus on the top 20% of the customer base, so we’re talking about the remaining 80%.
Let’s focus on investment products for an easier calculation, and let’s just evaluate the annual increase in AuM associated with new customers, shown in the following graph: in 90% of the simulated cases the increase in AuM is between 20% and 46%, with a median of 33%. The same increase is expected in commissions, if we assume that they do not change much by increasing the number of customers.
This result is affected by the assumption that all new customers are smaller than the current ones. All of them. This is an over-conservative hypothesis, since among the “small customers” there is a share of high potential customers3. By admitting that some new customers may be important in terms of size, the increase in AuM (and consequently in commissions) has improved significantly and 95% of the time is more than 40%, with an average of 80%.
As a whole, the results are comparable to the estimates of a completely different nature – for example, McKinsey estimates4 that advanced analytics lead to an increase in revenue between 15% and 60%.
This remains a “toy model”, good for having a first rough idea about the impact of digital innovation and data science. It is clear that there are other benefits of great importance: just think of the increase in the quality of customer service – something that will in turn increase customer loyalty and lead to an improvement in the life-time value.
The ability here is to find the right organizational model to combine:
- technology and use of digital channels – always available, with simultaneous parallel scalable processes;
- experience and professionalism of the consultant – only available at certain times, impossible to use in conjunction with and little/no scalable.
The benefit is an important leap forward in terms of quality, productivity and therefore profitability.
SideKYC® is an advanced data analytics software created by Virtual B for banks and insurance companies. SideKYC® can profile customers, identify individual needs and map them with the best product.
If you want to find out more, visit the page and contact us
1 – Wanted at Banks: Young Tech Pros with Old-Tech Smarts, fonte: American Banker
2 – Queueing Theory, autore: Ivo Adan e Jacques Resing
3 – Lo strano caso dei clienti mignon, autore: Raffaele Zenti
4 – Advanced analytics in asset management: Beyond the buzz, fonte: McKinsey & Company