BOTs by virtue of their design, cannot replace humans completely. Today, they are being looked at as a cost effective means to achieve efficiency in the system. Hence, a clear strategy on how to implement BOTs for increased customer centricity and efficiency is key to success, as it is not just about a machine answering some FAQs. BOTs by virtue of their design, cannot replace humans completely. Today, they are being looked at as a cost effective means to achieve efficiency in the system. Hence, a clear strategy on how to implement BOTs for increased customer centricity and efficiency is key to success, as it is not just about a machine answering some FAQs.
Most banks today are structured by business verticals with few functions that cut across as a horizontal, which is reflected in most touch points. Chatbots provide flexibility for banks to deploy them either horizontally or vertically. A hybrid approach can also be adopted.
It is also quite logical to assume that chatbots are in the continuous mode of learning and thus, practitioners of chatbots recommend that there should be hand-off from a chatbot to a staff of the bank, based on certain patterns observed while conversing with the chatbot.
Banks need to decide the best way to utilize bots effectively. These decisions will be driven by the banks aspirations, digital readiness and most importantly current and future customer profile.
Below are some suggestions for banks to look at in terms of making these decisions.
Vertical or Horizontal
A vertical bot would focus on purely banking function, while a horizontal bot would focus on multiple areas – perhaps including banking, shopping, travel and so on.
Banks need to decide whether they would like a horizontal play, with them playing the role of orchestrator and having access to the complete value chain or whether they would like to stay within the banking arena.
A horizontal play offers advantages such as access to extensive customer data and greater control of the value chain. On the flip side, it requires extensive ecosystem and AI capability across verticals, which banks do not possess today.
A horizontal play would work when a bank is a part of a conglomerate offering many other services. For example, Reliance Jio (which is a part of a larger Reliance Group) could offer a BOT which allows customer to shop from its retail stores, recharge its mobile bills, make payments from its payments bank and so on.
Having said that, our suggestion to banks is to start with a banking vertical-based bot, which can be built easily, and will play on their strengths. Once the BOT is established, it can be extended to the ecosystem to provide a seamless experience.
Pure AI or hybrid approach?
It is essential for banks to identify the processes that will be addressed by bots based on the level of complexity and the degree of automation required. Banks have the option of automating some very simple processes (for example, displaying balances or the last few transactions) on a completely NLP-based system. At the other end of the spectrum lies a hybrid, where a conversation starts with a bot but is handed over to a human later. For example, in the case of a customer dispute, a bot may gather the basic information, but hand it over to a human being for resolution.
If there is a fraudulent transaction on a consumer’s credit card, consumer would like to block the card first and as per human emotions, would like to get connected to another human to describe the situation and seek assistance- a clear case of hybrid approach.
A BOT at this moment might not be trained enough to spot a cross/up sell opportunity which clearly requires a skilled staff of bank to take it forward.
Hence, a holistic approach needs to be taken with a clear and deep understanding of all the existing processes.
We suggest that banks look at the low hanging fruit of pure AI to enable simple processes to start with. As the efficiency of the AI system improves, more complicated scenarios can be included and the relevant process hand-offs designed.
Identify Use Cases
As with most technology, Banks need to identify appropriate use cases with target segments in mind for chat bots. Where a bank would use a chat bot would be driven by the priorities driving the bank. For example, if streamlining internal processes is of high priority, bank may look at automating time consuming, repetitive tasks being performed by employees. For example, JPMorgan chase launched a bot to read and analyze complex legal documents, which saves over 360,000 hours of manpower.
If providing a digital, 24/7 customer service is a priority, then a messenger bot on a customer’s most used channel would be a good fit for use case. This is one area where banks have forayed into extensively, as can be seen in all the examples provided earlier.
The decision would be purely driven by the bank’s strategy, priority and readiness of the surround systems for “Botification”.
Bots have the capability to connect the dots and enable the banks to expand their ecosystem and provide a truly meaningful and profitable digital experience. Hence, we recommend that banks should start their journey with BOTs by identifying a good a business use case where they can provide the differentiation.
Quick movers have advantage in that their BOTs will start learning earlier than others, and will therefore evolve faster as well. Hence banks do not have the luxury of maintain a wait and a watch approach any more.