Creating the perfect chatbot

Kunal Patel

Council member at GLG

Views 314

Creating the perfect chatbot

08.06.2018 11:15 am

The world has gone chatbot crazy, certainly there is still a lot of focus on ‘virtual assistants’, as some people like to refer to them as. They are essentially used to mimic human speech where a human would be required via a messaging interface. As someone who closely works within the financial services industry, automation is certainly one of the priorities in large companies, esp. within the banking world.

It’s fair to say that the advent of chatbots has fuelled the drive in Artificial Intelligence (AI) as a whole in other areas. I have had conversations with a number of people across the financial services industry, there is a balance of those who are in favour of chatbots and those who are not. I came across some very good prototypes early on, they were quite basic but have evolved over time. Some organisations have already begun to or have rolled out services.

The Financial Brand have summarised a list of banking chatbots already in service – click here for more details.

Customers are the key to success

This seems an obvious thing to say I know, but the lines can be blurred if services don’t meet the need of the customer. After all, why invest so much time and resource when you’re not meeting or fulfilling your objective? It seems mad but believe me, it’s quite common place in industry. Chatbots should be fully focused on the customer experience, this means that they should be intuitive, naturally evolve and learn to become better with its responses. Above all else there should be a clear set of customer service objectives which can be measured by the company.

Conversable have an interesting proposition, their platform makes it possible for employees to write and modify responses for their bots and push automated responses live, without needing to know how to code. The console allows a user to deliver answers quickly, easily, and correctly, to their customers.

Take a look at Conversable’s chatbot building process here.

I really like the way they have broken down the process into six simple steps:

  1. Design the conversation 
  2. Build the conversation 
  3. System integration 
  4. Learning 
  5. Expansion 
  6. Advanced AI: 

The good, the bad and the ugly chatbots

There are some obvious pitfalls companies face when reacting to innovative services, a lot of it is just a lack of understanding of their objectives and implementation of the technology. I’ll try and explain what I mean by those two areas because this could be the difference between success and failure.

Be clear on the objectives

Customer experience! Customer experience! Customer experience! 

Don’t be lead by what others are doing but think about areas where you need to improve and what pain points you are experiencing. The driver here is to think about improving efficiencies, service as a whole and how you’re best positioned to deal with customer experience.

Don’t look at chatbots as quick fix to your problems, so you can avoid customers and just focus on your revenues.

Holding onto customers has now become one of the biggest challenges across many industries.

Half-baked chatbot product

If you’re not clear on objectives the consequence will be a half-baked product or one which is overloaded with functionality which tries to cover all bases. This will no doubt compromise the quality of the customer experience, and as we know this is the main objective. You can see how this will end I am sure, the consequences could be disastrous given the competitive nature of our current digital world.

The other direct driver is the perception of chatbots and their use based on the plethora of services which are currently out there. The more half-baked bots out there in industry only compounds people’s perception. As someone who focuses on product as a day job, this annoys the hell out of me.

The current customer crop, segmented into millennials, generation Z and the up and coming generation alpha have little time for bad products. If it doesn’t work first time or the experience is poor, it’s quite literally toast and the recovery from that isn’t worth thinking about.

It really boils down to some simple thinking and the following steps.

Be clear on your objectives, have a concise business case in place, understand the customer journey intimately and have the courage to make changes where applicable. There’s no excuse not to put the customer first and focus on the customer experience, regardless of AI and machine learning technology being in place.

 

Latest blogs

Javid Khan Pulsant

Using Cloud to Relieve the Compliance Burden

A recent survey of more than 360 enterprises revealed that 86% are dealing with the complexity of multiple types of data and/or data-related processes subject to privacy and security compliance requirements. Just 61% say that their organisations are Read more »

Lina Adolf-Orup Fingerprints

Fingerprint on the Pulse: Biometric Payment News

It’s quite hard to believe that we’re already entering Q4. For the biometrics industry, it’s been a unique and exciting year – one marked largely by its increasing convergence with the world of payments. For payments and biometric lovers alike, it Read more »

John Bertrand Cognizant

New Technologies create new opportunities in trade finance and working capital

At Sibos 2014 in Boston blockchain was the talk of the show and the movement towards adoption of new technologies was up and running. As Sibos 2018 opens in Sydney, APIs and open banking based on their adoption will be high on the agenda, indicating Read more »

Marten Nelson Token

PSD2: The real RTS deadline is closer than banks think

Let’s work backwards. Most banks know that the final deadline to comply with PSD2’s Regulatory Technical Standard (RTS) is 14th September 2019. Eleven months away.  Following the amendments to the RTS, however (based on industry consultation and Read more »

Alexander Peschkoff Trusted

Brex - not Brexit - is the shape of things to come

There are a lot of discussions on LinkedIn about the Brex duo - two 22-year olds who built a unicorn in less than two years. “Young and inexperienced, did not invent anything new, there are thousands of similar products, etc”. Well, it’s not that Read more »

Magazine
ALL
Free Newsletter Sign-up
+44 (0) 208 819 32 53 +44 (0) 173 261 71 47
Download Our Mobile App