Improving commercial strategy of banks, all thanks to Artificial Intelligence and OCR

Ver versión en Español

Inteligencia Artificial

A

I (artificial intelligence) undoubtedly offers a considerable number of attractions to the financial sector.  Thanks to AI banks are able to make their human teams much more efficient, it allows them to focus on high-value tasks while technology automates the rest.  Thus, they are able to increase productivity by offering better quality customer service.  Likewise, automated document processing solutions, using AI and machine learning, not only alleviate work for offices and call centers, but also analyze operational risks and predict customer behavior, detecting commercial opportunities.  This is one of the most interesting and useful aspects of this revolution.

A significant percentage of organizations choose to rely on an external provider to build their personalization strategy offer through AI.  This personalization involves monitoring the various channels and contact points to identify preferences and shape impacts.  We are used to music or video-on-demand platforms recommending content, without realizing that banks are using similar techniques.  For example, when our bank’s application offers us the possibility of adding the accounts of other entities, we are allowing that bank to know our habits and financial situation more precisely and in real time.  This powerful information will be used to draw up profiles and, from there, open the door to a loan or mortgage, send warnings and draw up hyper-personalized recommendations.

Document processing platforms based on artificial intelligence can, for example, scan in seconds the transcription of one of the thousands of calls received daily by call centers.  These calls, without a tool capable of filtering what is important and what is not, are nothing more than very heavy and worthless data files.  The automated screening process, however, makes it possible to respond immediately to the customer’s intentions, such as unsubscribing or showing an interest in purchasing a new product.  It is much more difficult to find a new customer than to retain a dissatisfied one, so implementing this data mining is not a trivial matter.  These solutions are not simple search engines; they learn, they build models and they use what they already know to apply it to another task, in a process known as transfer of learning.

Banking is moving towards proactive and emotionally sensitive digitalization so that the customer relationship grows and improves, there is however plenty of room for growth.  According to the latest World Retail Banking Report prepared by the consulting firm Capgemini and the European Financial Management Association (EFMA), 75% of customers are attracted by the new possibilities set off by digital banking.  Forty-nine percent think that the relationship with their bank is not satisfactory, and a similar percentage wish for a more “emotional” connection with their bank.  The financial sector should know how to manage this discontent, prioritize the user and promote an environment that exploits the advantages of digitalization, correcting traditional mistakes such as the lack of personalization or transparency.  However, 70% of banking executives are “concerned” that they are not up to the task in terms of data management.  In fact, up to 95% acknowledge that their systems are obsolete.

Fintechs have a unique opportunity to find clues within all this unstructured data to refine service in every customer-entity interaction.  By finding traces of satisfaction or dissatisfaction in language, AI can decipher feelings and, from there, meet their needs and follow up, without losing sight of the risk level.  It could pinpoint which products are garnering the worst feedback, compare thousands of customer contacts across channels and identify the most common problems.  From there, deliver structured data to make decisions, segment user groups by financial profile or level of satisfaction and, for example, develop an ad hoc marketing campaign.

Automating document processing significantly reduces errors and administrative costs and is an essential support for ensuring regulatory compliance.  However, AI would take care of the most cumbersome part, but leave the most sensitive operations to employees.  Personal attention will not disappear, but it will become increasingly specialized.