Chatbots and the post-app era: Implications for banking

This post was prepared in cooperation with Radek Jezbera, a great friend and good professional in financial services, management operations models and financial services innovation.


We have now experienced the first 16 years of the 21st century by now. Eventually, the pace of radical innovations has accelerated again and we see the society transforming very rapidly. The way we interact with each other, the way how we consume services and our expectations have shifted towards immediate fulfilment, instant gratification.

It all began with the internet which at first had a form of a bulletin services accessed through a text console interface by scientists and computer geeks. Then, we have gone through the phase of internet browsers on PCs while we have witnessed first steps in e-commerce and emerging online banking. We are now in the third major evolution stage as we consume information services on our smartphones mainly through a number of single purpose apps which are pushing the built-in browser in your smartphone into oblivion.

It seems that there is another trend emerging in the “smartphone stage” which may become bigger than smartphones themselves. The trend is a fundamental shift from a user interface built around shape and form of the device e.g. a PC web browser built to be handled by keyboard and mouse to a human centred interface e.g. interacting with fingers, gestures, voice, or perhaps in the future with thoughts (natural user interface — feedback). But why we would need apps for human centred interactions with technology? Apps will transform into services that will work silently in the background while we will communicate with artificially intelligent beings whose ancestors, well alive today, are called digital assistants, chatbots, Siri, Alexa or Cortana. They live on a cloud platform and manifest themselves when we push the button on our iPhones, when we address them in our living room, or most commonly when we are typing or talking to our favourite messaging app.

The messaging apps seem to become the bridge between today and tomorrow. They are the most frequently used apps today. Their increasing usefulness is the number one cause for the mass extinction of the single purpose apps. There are becoming platforms themselves enveloping services which in the past we needed to download, launch, and register within a separate app. For illustration, the most popular Chinese messaging app WeChat apart from communicating with friends incorporates food delivery, ordering a taxi, arranging a date, disseminating news, sending a payment, asking for a credit, and more to come. All these useful services are accessed by typing or dictating text to your chat buddy who is in ever increasing number of cases an intelligent robot rather than human.

There is another reason for the emergence of robots and conversational services. The advent of massive computational complexes and ultra-fast processing routines allow storing gazillion times more on the web, as well as capture near infinite new trove of data. Originally we have worked with a simple index of the data that allowed us to query the data with simple requests. These requests retrieved data back to us based on the query. However, today, the paradigm has changed.

New computing capabilities and natural language processing allow computers to grasp and contextualise data — as well as put it to new uses. When in the past we searched for data, now we look for meaning and context.


You, as a bank customer, have to overcome several hurdles when in need to engage with your bank.

Firstly, you do not want to go to the branch. Your nearest branch was probably made redundant already as a part of banks cost savings, still reeling from overexposure to mortgages. If it is not the case, the opening hours have been shortened and therefore you will face the queue during the peak hours which are coincidentally the most convenient time for you to visit.

Secondly, you do not wish to talk to a call centre agent as you correctly expect rather long waiting time on the line, awkward authentication questions, and then difficulty to explain the issue on the phone.

What is left is the information you can get on the website or within the web / mobile banking app. Fortunately, a number of banks have introduced “a chat with the agent” feature within their online banking offerings. The trouble with that is that the response time is sometimes unpredictable as the agents get busy.

When they get to you they are often too distracted to understand more complex issues as they are chatting with multiple clients at the same time. Although the “live chat” feature seems to be the most promising among other options, the issue lays in the fact that bank let you chat with them only within their own app. The mobile banking penetration in the UK in 2015 was just below 30%.

The reason for this non-impressive figure lays in the three obstacles:

  • Requiring client to download the app
  • Getting client used to interface and the way how to use the app
  • Client forgetting about the app on her smartphone

Despite adoption hurdles and lack of functionality, you don’t want to fall back for the physical solution where online ones are present and self-evident. If you can chat with your friends, search for information and have it found, your problems, be that a recipe, a food order or other menial task resolved online, why should you seek another cumbersome way of performing the task?

It is becoming obvious that although built-in messengers with live agents are the way in the right direction, they face fundamental obstacles given the current set-up of the app economy. A solution for the post-app and, potentially, a post-smartphone world is needed.


Interacting with a 24/7 availability intelligent robot (chatbot) on a commonly used messaging platform overcomes all these hurdles. There are no conversion hurdles in downloading a specialised app or log-ins at the bank’s website as client already has her favourite messaging app on the phone. The app is most likely located in a prominent place on client’s smartphone being opened several times a day.

More importantly, instead of navigating menus while searching for the transaction, the client just types or says what she wants to do. The chatbot on the other side will understand the command. Sometimes, it will offer a simple menu of options to ease the interaction with the client. A good example is a recently launched TransferWise bot at Facebook’s messenger platform, who offers a list of currencies and previously saved payment recipients. Many others chatbots are already helping people to track prices for plane tickets, remittance services quotes, order food, order a manicure, etc.

With chatbots in financial services space, security should be taken very seriously. At the moment, the client is asked to log-in with their internet banking credentials to authorise themselves. The communication itself is encrypted by default. A number of messaging platforms offer a type of encryption that makes impossible to Facebook or others to read the conversation between the bank and the client.


Chatbots in numerous visible and invisible forms will take over many tasks which are today performed in apps or at websites. They are the very foundation of the coming new era of post app store / post-app economy.

Their initial success is based on their ample advantages in removing usage hurdles, speed, availability, and simplicity. Their success in the future will be driven by tighter integration with existing authentication mechanisms and further improvements in NLP (Natural Language Processing).

There are vast opportunities for banks in the chatbots space ranging from obvious ones like customer service and initiating payments to more sophisticated uses like personal finance management or collecting late payments.