Millions of small entrepreneurs want a loan at the click of a button, but banks don't know it, says Mirek Kolář from Adastra Business Consulting
Nowadays, technology makes onboarding much easier for customers across segments. Onboarding is the process by which they become part of the customer base of a company so that they can agree to receive its services or products. It also brings many benefits to companies. It offers them opportunities to comprehensively understand onboarding – as a process during which, based on data they collected, they can help the customer, get to know them, find out more about them and offer other services or products tailored to their needs. All of this takes a matter of seconds, minutes at most, and helps achieve the best possible customer experience.
The advent of the pandemic has intensified customer demand for convenient onboarding. They expect to handle it with a single app, from any device, and in only a few seconds. Regardless of the segment, they want to negotiate for what they want to purchase or what product or service they want.
Until two years ago, a number of surveys, such as the one by PwC, showed that three quarters of companies were already offering online onboarding using a web or mobile app. Research also revealed that many of these companies did not have smart assistants or advanced systems to handle text conversion and fast document processing.
An Experian survey then substantiated this insight with the responses of 380 top managers and leaders, as many as 70% of whom admitted that their companies did not provide an optimal customer experience across customer life cycle contact points.
At the same time, even two years ago it was possible to onboard customers relatively smoothly. Today, technology has taken another leap forward and is helping companies to forge and build relationships with their customers - relationships that show customers that the company knows them based on data. This is a relationship and approach that will also make customers want to come back.
This is why it is good to take a holistic approach to onboarding – using access and the latest technologies to make it as user-friendly as possible. And to help the company achieve a better relationship, better data and more efficient sales without bothering the customer. As for us, we use artificial intelligence.
Remote onboarding, during which the customer does not have to visit a brick-and-mortar branch of the company, is a necessity at the time of Covid. In some sectors, however, we have long perceived the adoption of onboarding technologies to be a far more uneven path. One of them is the financial sector. According to a Deloitte survey, at least 38% of customers leave the onboarding process of traditional financial institutions during that process. This is often due to frustration with the number of necessary steps and paperwork or the ultimate need to visit a branch.
This type of onboarding requires identification, and possibly the submission of other documents, which then need to be checked, and the manual entry of the necessary data into systems. This takes time and a lot of manual work, and data entry often leads to errors. Systems based on optical character recognition (OCR) and natural language processing (NLP) take over these processes, and process documents immediately. They convert documents into digital form, and we know from our own experience that this reduces the costs associated with manual document processing by up to 80% and time by up to 98%.
Based on our implementation, we know that it only takes seconds to process documents. The system then verifies the obtained values in public databases (address, name) and against each other (consistency of read values). The output is data in a machine-readable format, which is ready for further processing. Likewise, the data mining and verification system aids compliance with regulations requiring biometric control using a standard web or mobile camera.
However, some types of companies encounter fraudulent behaviour during onboarding, despite supervisory processes with defined rules. These include, for example, telecommunications companies that deal with attempted fraud in connection with subscriptions and tariffs (customers who do not pay for services and equipment). Artificial intelligence helps with this type of business problem, too.
Smart machine learning-based algorithms detect attempted fraud in near-real-time. Thanks to constant learning, they are also able to detect the latest, sophisticated tactics of fraudsters based on data. The deployment of this type of solution allows companies to minimise the number of such cases, improve risk management and strengthen the fraud prevention process, in our experience by up to about 30%. At the same time, technology helps reduce the number of false positives by up to 60%.
Another level of control and a recent trend is the behavioural detection of fraud when filling out, for example, online questionnaires and documents. The AI uses mouse movement, click speed, keyboard typing, number of typographical errors, and other metrics in behavioural biometrics as input data for analysis. Technology can thus identify an individual person and the general pattern of normal behaviour, and therefore deviations from it.
Whatever the customer is looking for in a company, they will often appreciate a virtual assistant or a chatbot that helps them with a given task when onboarding or filling in more complex documents, such as investment questionnaires.
Chatbots and assistants, which have been experiencing a major boom in the last three years, operate on the basis of natural language processing. They are able to interact and evaluate what is being asked by the user and compile and send a relevant answer. This is possible thanks to the connection between the assistant/chatbot layer and the understanding layer, for which quality data, from which the assistant learns, is essential.
This helps the app to decide whether it is necessary to ask other systems to answer a question and what answer to send to the customer. In addition, many customers prefer them to seeking help from customer care operators – the technology works at all hours of the day and responds immediately.
Within a few seconds, a company can automatically obtain its first verified data and create an overall input image of the customer, to which further data can be added. The more data, the better the image companies can have, at the same time aiding more effective decision-making in the overall context.
At the same time, the data is used to modify and improve the digital experience with the company in question. Thanks to constantly learning algorithms, this is a continuous process. Artificial intelligence, specifically machine learning, can easily find new relationships and rules in customer data, and the company can also significantly improve its offering.
It always pays to have a holistic view of data, and in the case of customers, these are the so-called 360 customers. On online and offline data can be used as the basis of an image for the subsequent creation of personalised interactions during the customer life cycle, such as the acquisition of property (car, real estate) or life events (wedding, addition to the family). It is therefore easy for a company to offers the customer the consolidation of a loan or mortgage when they need it, through the most suitable channel for them, and at a price that they will be willing to accept. Moreover, it is in the financial and insurance sectors that the use of artificial intelligence in bidding is expected to be essential.
In addition to purely business purposes, data is used to adapt content according to communication channels: personalisation of web content, in the client zone, annual newsletter, etc.
In addition to tailor-made offers, artificial intelligence also significantly helps with upselling (an approach in which sales offers the customer more expensive versions of goods or services, or upgrades) and cross-selling (techniques for selling related goods or services that complement those already available). This happens at the moment when the customer is purchasing specific services and products or access to the company.
Thanks to the company's customer data from all the different channels, contracts and histories, vendors know which other products and services they can offer so that they are relevant for the customer, and at the same time for the price that they will be willing to pay, or offer such that complement their existing products and services. In this case, artificial intelligence easily calculates the optimal highest possible price that the customer will be willing to pay, and easily recommends products and services at the appropriate time and tailored to the specific customer.
There are a number of options and processes for deploying artificial intelligence that really help. Therefore, their adoption and implementation should be considered in terms of need, business goals, and involvement in the operation of the company and its processes. A proven innovation framework exists to guide companies through the entire process, from idea to implementation. At the same time, the AI Open Days, in which we help companies generate business cases for the use of AI tailored to their needs, have proved their worth.
This article was written by Ondrej Vanek, the co-founder and CEO of Blindspot Solutions. It was published by IT Systems 3/2021. You can also read it in Czech at systemonline.cz.
We will contact you as soon as possible.