Il ya quelques semaines, nous annoncions l’arrivée prochaine des “clients-robots”, soit le développement d’assistants virtuels utilisant les ressources de l’IA générative et capable de négocier à notre place, dans les meilleures conditions, nos chats des goods and services.
We emphasize that banks will be especially concerned about these new assistants, capable of replacing them in their traditional mission of advising clients. It is therefore not surprising that one bank, the Commonwealth Bank of Australia (CBA), commonly recognized as one of the most innovative in the world, has decided to reverse the problem by developing “synthetic clients” to develop and test its products.
It is about defining virtual customers that imitate human behavior, to better understand the expectations and, especially, the reactions of real human customers to new products. One step beyond person In the field of marketing, which are essentially typical customer profiles, generative AI, with its ability to mobilize and model countless data, should help to “manufacture” real virtual personalities, to better understand the possible reactions of real customers.
Generative AI to improve customer responses
To do this, CBA algorithms will be “trained” on simulated experiences of daily life and fed by analysis of customer reactions that chatbots are now able to collect. This way we can simulate the different types of searches that customers may have performed before they were offered an offer. We will also be able to better gauge reactions in difficult contexts, such as those of customers who have suffered fraud or who are grieving. And generative AI will continue to be used to make messages sent to customers more effective.
Let’s face it, there’s something a little fanciful about these ads! As if we expected wonders from AI. As if, finally!, marketing managed to realize its greatest fantasy: capture “the” customer, put it in a test tube and study it, dissect it to its liking in the laboratory.
The process is still in its early stages and it is difficult to predict its future results. Of course, the potential obstacles are easy to imagine: spending a lot of time and money to learn what simple common sense would have been enough to understand; not to better identify real customers, but to calibrate them in relation to “synthetic” customers, quite artificial models (isn’t this already the frequent error of the typologies used by marketing?).
Human relations, soon the exception?
In any case, for the CBA its approach must translate into providing assistance to its advisors in relation to their real clients. It’s about increasing customer relationships and not automating them. From this point of view, we can consider that generative AI will be able to provide useful recommendations to advisors in certain situations and that it will contribute, more generally, to hue relationships by multiplying the factors taken into account.
But what if tomorrow real clients were accompanied by virtual assistants or even replaced by them in certain cases? It would then be necessary to introduce largely automated hybrid clients between the synthetic clients and thus we would have robots that would imitate the behavior of other robots. What if tomorrow, in banking, human relationships no longer represent the rule but the exception?
Source: BFM TV
