“Our goal is to teach a bank to use the opinion of social networks”
—Could you please tell us: what is Innodata’s business and how can it help a bank to understand the attitudes of its customers using their publications on social networks?
—One of the key products created by our developers literally from scratch is a system for monitoring and analyzing the content of social networks. This system has been designed and operates to help banks that communicate with a large number of customers. These are the banks that serve individuals, that is, retail credit organizations. These are also the banks that provide services to legal entities - in the event that the bank maintains accounts of individual entrepreneurs, small- and medium-sized businesses. In other words, these are nearly all the banks in Russia. Any bank is in contact with a large number of customers who post reviews about the credit institution’s activities in social networks. The number of reviews posted by customers is huge. For a bank with a large customer base, the opinion about its services is sometimes the key to determining whether its chosen strategy is correct. Customer reviews are needed to understand the quality of products and the ways to improve services offered to customers.
—Why did you start monitoring customer reviews on social networks? After all, customers publish reviews about a bank’s service in very many places, for example, on banking forums.
—Even before social networks became an integral part of our life, many departments within a bank were engaged in monitoring the customers’ opinions about it. Banks always employed and keep employing marketing experts who conduct service quality surveys among the customers. There are PR services that are involved in promoting a bank in various media. Today the market as a whole has changed considerably. In our age of breakthrough technologies, social media are playing an increasingly important role. Of course, the bank’s support and service assessment services persist, but the tools for their operations are somewhat changing. In fact, social networks are now upgrading many tools that a banking institution previously needed to assess the attitudes towards itself. The bank previously had to look for its customer, run after her/him, ask her/his opinion, and collect reviews.
—Banks are still doing it: they make calls, send letters, conduct polls, intrusively offer something ... If a bank arranges its activity with a customer in a social network in a correct manner, it will have an excellent tool for getting high-quality feedback from customers at its disposal.
— Of course, banks now also run after a customer and ask his opinion. However, a testing ground for studying public opinions such as social networks is now available. If a bank arranges its activity with a customer in a social network in a correct manner, it will have an excellent tool for getting high-quality feedback from customers at its disposal. In addition, the bank will have a correct understanding of the customer’s attitude to it. Our system has been developed and aims at helping the bank – not to replace any important processes within the bank itself, but literally to provide help to the PR service, marketing service, and quality control department.
— How does your system work?
— The system comprises a number of modules. One of them is specifically designed for working with the PR service. This module has been developed to identify newsworthy events. For example, a bank launches a new product on the market, it has in pipeline a serious deal with a partner. This is the original material about which the bank structures are notified in advance. For example, a bank launched an advertising campaign for its new product on a TV channel. People immediately start discussing this ad on social networks: “Well done!” or “What bad conditions they offer!” Accordingly, the responsible departments of the bank will be able to immediately see the reaction of viewers, social network users, to this event. With our help, bank employees will know in advance where this reaction can be monitored. Also, using our system, they can assess the effect of the information event: positive, negative, or neutral. They will be able to trace who writes what and in what terms, and whether the tonality of posts about the service is positive or negative. They can identify the users who are especially negative, and those who are delighted - it sometimes happens like this! There are also cases when an event happens in the bank unexpectedly and an instant reaction is needed. Should this be the case, getting timely feedback from customers through social networks is of great help, because the financial well-being of a banking institution directly depends on its reputation.
— It would be relevant to recollect that at the end of last July two backup cables of the data center of a large bank were interrupted by an excavator bucket on the Garden Ring ...
— Yes, a serious technical failure is the most unexpected thing for a bank. Everyone on social networks, whether customers of this bank or just passers-by, immediately begin to discuss the event. What is the tonality of the discussion? Someone begins to curse support and technical services; someone shows understanding. Anyway, people instantly assess the bank’s reaction to the event in social networks, for example, whether the bank warned its customers by phone or a message. Customers assess what they are told in the call center. In fact, the bank’s top managers immediately obtain a transparent picture of what people think about the bank and its activity in an emergency situation. We had a case in another company, not a bank, with which we cooperate on the basis of our system. The company’s CEO learned about the incident before his subordinates in the field knew about it. He called them and asked: “What has happened there? Why are people indignant?” The subordinates answered: “Everything is OK!” This implies that, according to the systems and the employees’ reports on the spot, the company is doing well. However, in fact, people on social networks were already actively writing that very serious problems had occurred, and it was high time for the company to start responding to them.
Accordingly, our system also provides a kind of PR; it is geared towards identifying newsworthy events and customers’ reactions to unexpected events.
—Why does the bank need to monitor the reaction if events have already occurred, and all that the bank can do is to fix the breakdown, eliminate the cause of the failure, and just apologize?
— It is needed to ensure that the reaction to events is correct and under control. Our system helps the bank identify which sites react to the event and in what way. It happens that this negative attitude is spread not by live customers but by accounts specially created in advance, which are designed to react negatively on everything that happens in the company. It is very helpful to know in advance who instigates the major groups of people. This knowledge also helps correctly respond to negative events in a banking institution. Our system has an important module, customer care, which enables handling customers in a correct way.
— What does your system change in the way the bank handles customer requests?
— What procedure was used to handle customer requests or complaints? It was quite usual that a customer has to submit a complaint in writing. It was necessary because the petitioner had to describe the problem, exhibit her/his attitude to it, and formulate the claim in a structured way. The complaint was then reviewed by the company. Nowadays, in the digital age, all companies introduced electronic filing of requests. It was sufficient to upload a request on a website or on some specialized forum. Nevertheless, it was necessary to perform an operation on a remote site. To do so, one had to find the site, log into it, register, upload some kind of application, and then wait for an answer. Sometimes the waiting time was very long.
— What changed?
— Now people moved on to the next stage. They no longer want to log in any system or in forums; they just want to post a complaint on their page on a social network like, for example, “I was poorly served at the office”. People today expect that the bank identifies the customer, the bank employee independently finds out about her/his problem on the social network and in fact creates the customer’s application to the bank herself/himself. Then she/he has to resolve the problem and return to the customer either with information about the measures that were taken or – even better – with a compensation. The bank’s response depends on the case, severity of the event, and the importance of the customer for this bank. However, it is already an accomplished fact that people are waiting for an instant reaction to their posts on social networks. Moreover, now, if a person does not like something, the first thing she/he does is to publish her/his opinion about the bank on social networks in the form of a post.
— Why is it so?
— The reason is that the person wants her/his friends to be aware of her/his problem and disadvantages of the bank, and to make appropriate conclusions. The only correct approach for any company, including banks, is to learn how to respond properly to such negative messages. The complaint should actually be perceived if not as a gift, then as an opportunity to improve the service.
— And let me ask the same question again - how does your system helps the bank in this issue?
— Our system enables rapid identification of such messages. We scan, roughly speaking, 95% plus of all messages in all popular social networks, and enable a reaction to a negative message with a delay of no more than a minute, that is, immediately. The system makes it possible to do this, even if the person has not sent a direct appeal to the bank’s site, that is, failed to type in the @ character and the company’s hashtag. The customer just wanted to complain and vent her/his negative attitude. If she/he only mentioned the company, we are able to submit such reviews to bank employees. Next, employees can process this appeal in a correct way, that is, get all available information: who is the author of the post, how many subscribers she/he has, and how popular this person is on the network. This implies that we automatically find the person in the company’s customer base, and the bank employee gets not only the person’s history on social networks, but also the history of her/his relationship with the banking institution. Bank employees can understand in this way how valuable the customer is, whether she/he uses any bank products, the number of these products, for how long she/he is the bank’s customer, etc. Thus, having obtained the entire set of customer data, employees of the PR-service, marketing department, SMM who directly handle customers in social networks, can create a customer request themselves and send it to the bank’s internal services for analysis. Otherwise, they can contact the customer to promptly give her/him a kind of feedback: “We received your appeal and we will consider ASAP.” Practice shows that such a behavior – when the bank shows a proactive attitude to any dissatisfaction, immediately gets in touch with the customer, and demonstrates how valuable each customer is for the bank – makes a very serious positive impression on customers.
— People mostly write about their negative experience in communicating with the bank. Are you not afraid that your system will only collect negative reviews? After all, people simply do not have time for positive feedbacks, and often they have no desire to do it.
— We are looking at the total number of messages about the bank that have been found. In principle, both positive and negative reviews are collected. A company oftentimes spends a lot of money on mystery shoppers, on some kind of verification by people from outside the company. However, a customer on a social network speaks directly about a problem. No secret purchases are needed. Everything is already in place, everything is available.
— Can your system work not with the general data set of opinions referring not to the bank as a whole, but to some specific product?
— Our system can collect any mentions about individual products. Moreover, we can see both references to a specific bank, and we can configure some references to products. For example, if a bank offers loans under the military mortgage program, it is possible to collect the attitude of people in general to such a term as “military mortgage” and everything that is related to this term in the minds of potential customers. This can be done to understand whether people comprehend this term correctly or the product should be renamed.
— Do you handle requests from customers in difficult life situations?
— There are such messages when a person is in a difficult situation and urgently needs help. Such messages should be escalated so that the bank’s specialists could react to them immediately. There are also messages where a person simply complains about her/his hard lot, and such an appeal should be considered according to a regular procedure. Having collected all this information, processed it, distributed it into categories, the system either transfers the entire data set for a kind of manual analysis to employees who handle messages, or, for example, automatically sends it to the backend system to have these requests processed.
— Will you have a specially trained robot to respond to customer complaints? It’s so trendy now!
— In the nearest future, we’ll release additional functionality for our system: it is a kind of built-in chatbot that answers some basic questions and solves trivial problems. For example, a customer submits a specific complaint, but the system lacks some information about her/him or about her/his problem, for example, the customer’s full name or the bank office name. The system can automatically ask a person in a personal message: “Could you, please, specify the contract number? We would then be able to promptly respond to you.” In addition, the bot will prompt and recommend certain questions that the operator still need to ask.
— Well. You outlined which banks would need you and your system. But I can’t help but ask - have you already offered your product to banks, and do you have any feedback from them? Which direction are you heading? Are banks interested in this product?
— Indeed, we are now trying to actively promote our system on the banking market. We have several agreements on pilot projects. We are already implementing them.
— What is the lead time of entering the market for the pilot projects? When can you tell about it?
— I think the run-in will take place in the next two or three months; I believe that it will happen by the fall.
— What are your system’s advantages and differences from competitors?
— Current players in the marketplace that specialize in such systems often offer SaaS solutions, that is, “service as a service”. Indeed, this is enough for many banks, and it is sufficient to start operating on social networks. However, we already gained a successful experience in combining information that we collect about customers with the bank’s customers in order to integrate data collection from social networks with the process of handling requests used in the bank, or create it if it is not available. This is actually a very important issue, because often - and in some companies we encounter this phenomenon - social networks are handled, and this is the responsibility of one department, and appeals are handled as well, but this is the responsibility of another department. When a response is sent to social networks or via e-mail, this implies that different bank employees respond to different bank customers. That is why bank specialists are often unable to reach a common understanding and answer the same questions. Therefore, we believe that it is of importance to integrate our product into the bank’s infrastructure. We have experience, and we position our product based on it. It integrates with completely different solutions, be it a proprietary system designed by the bank’s developers or some ready-made enterprise products like CRM Siebel, Microsoft, or SAP. We have experience in coordinating with various solutions. Therefore, perhaps, our advantage is precisely that we propose not to separate handling social networks into a specialized service, but embed social networks into the bank’s standard processes. Should this be the case, the effect that we were talking about will occur.
— Does your system help predicting customer preferences?
— We are now developing a new module intended for conducting marketing research in banks. This program module, in addition to the direct tracking of what is occurring in the company itself, will make it possible to carry out instead of, for example, conventional marketing surveys and market analysis, an analysis of which of competitors’ products are more attractive to customers or, on the contrary, less popular, and compare them. Using aviation industry as an example, we consider air travels and, in particular, take the Moscow - London flight, and compare different airlines with each other. Accordingly, if we fly with the same airlines, then passengers like something more and something less. For example, we better like the quality of meals and seats and the entertainment offered by a specific airline. Accordingly, some products in the banking industry can be compared in about the same way. For example, offers differ in the case of cash loans, and this is a fairly universal product. Therefore, it is possible to track where a larger amount can be borrowed, where the interest for issuing a loan is lower, and where it is easier to pay off the debt. Based on the information collected, the banking institution can develop its strategy.
