Customer Service Chatbots - Hype or what?

Through this article, I am expressing my thoughts about the growth factors, key market observations, technology stack and current hype around customer support chatbots.

I am mentioning Key driving factors for the growth of chatbots -

• Mobile Messaging scale and enterprise messaging growth. Third-party mobile apps(Facebook, Kik, Skype, Line, Telegram) are rapidly heading towards 4B active users globally. They are going to surpass the number of active users on native mobile applications. There is a trend of Messaging-as-OS where customers prefer to talk over messaging rather than calling. 
• Consumer’s reluctance to install apps and availability of new interactive interfacing model with online services. 
• Following the model of Asian messaging apps WeChat and LINE, the owners of other messaging apps want to turn them into platforms. Facebook, Slack, and Microsoft have all opened up their platform for developers. Facebook hosts more than 30,000 chatbots on its Messenger platform. 
• Customer service solutions are conversational in nature. Customer support is the most resource intensive department. 1990-2005 saw emergence and utilization of BPO industry to manage calls. Contact centers have evolved over the last 15 years, with rising popularity of digital communication channels. Digital channels(SMS, live chat, social media, email, text) account for more than 35% of inbound inquiries.

Market Observations:

• Global chatbot market has seen rapid growth in recent years due to development in AI, digitalization of customer service channels, technology developments in NLP and trend of the messaging-as-OS platform. 2017 is being called the year for conversational commerce. 
• Bot development has changed since mid-2000’s when virtual assistants in live chat were the only automation in this market. Customer questions were answered from a pre-determined directory of responses. These responses were out of context and unhelpful. Studying these interactions led to advancements in Natural Language Understanding(NLU) which aims to identify the intent behind the incoming inquiry. Paired with AI, today’s technology is helping to put “thinking” power in bots, rather than feeding pre-assigned answers.
• The market for chatbots is very nascent and highly fragmented, due to the availability of open source NLP engines. 
• Broadly, there are 2 types of Chatbots – Decision tree based bots and NLP(ML) based Bots. ML-based bots require a huge amount of right datasets to provide a thinking ability in bots. Even major players are not able to solve the underlying tech stack. Giants like Apple and Amazon are working on NLP and ML based bots. Siri and Alexa can also handle only simple, structured requests. The capabilities of scripted bots are limited to the product designer’s ability to imagine all the possible scenarios and the response iterations in the databases. The failure rate is high in scripted bots. Most of the bots today are decision tree based bots, which undervalue this market. Due to the hype created around chatbots and poor ROI, Gartner has put chatbots in ”Trough of Disillusionment”
• According to Tracxn, Globally, More than 50% of the companies founded in 2016, are trying to automate customer service messaging using AI and chatbots. Out of this 70% companies are offering isolated chatbots, which will eventually fail to provide ROI. Chatbots need to analyze every question from every customer from every channel and use that data to continuously train a "centralized system of record". 
• Bigger and legacy players will have more advantage since their ML algorithms have been trained on the huge amount of datasets, and NLP engine refined with different types of queries. Bigger players like Aspect Software, Nuance, Interactions, Next IT, Inbenta are the leading players in this market

Tech Stack:

• Chatbots must need to ensure that they have the scale, intelligence, and technology to handle specific tasks. The biggest challenge has been curating a useful database of information from they which can work. The aim of the upcoming companies like Digital Genius is to develop a chatbot that learns from the dynamic data of a contact center platform. A good chatbot crawls for any kind of information available in the contact center platform, including customer files, past interactions, FAQ, product knowledge, and more. Whenever the data is updated (for example, after a customer service agent closes out a support ticket), the information is crawled again, and the chatbot’s knowledge base stays updated.
• For a good chatbot to be developed and true ROI to be achieved from automation, the repetitive back-end(high volume, low end) work of customer service reps need to be automated. Chatbot needs to have Robotic process automation capability.
• There is a need for both custom and standard systems integration. Custom integrations are required with companies internal and proprietary database. Standard integrations with popular CRM, knowledge base, customer support software, and more, to make sure bots complement brand’s existing business processes.

Market sizing:

• According to industry sources, Customer service software market is $20B.
• Phone calls account for 65% of Interactions while digital channels account for $35%. Digital customer market comes out to be: $7B
• Chatbots(Automation) usage is limited only to Enterprises and has not kicked off. Assuming 15% of software spending is done on automating digital channels.
• Chatbots(messaging) market size comes out to be - $1.05B Globally

If you have any suggestions or queries, you may comment here or write to me - [email protected].