The age of ChatBots
Posted by admin on June 29, 2017
This is the era of text messaging; people prefer to send text messages rather than talk on phones, as that gives them the freedom of doing so many other things at the same time. This lead to creation of so many text messaging applications, which allow you to talk one to one or interact in a group. Which in turn has given rise to a new avenue for businesses to explore; can this messaging "addiction" be leveraged by businesses to make that sale? The conversation (Chatting) with customers with humans behind the desk has been done for quite sometime, remember Call centers, but scaling it to the level where your business is ready to chat with each and every customer on your website is a different ball game all together. That's where ChatBots are being proposed as the solution.
Chatbot is not a new phenomenon, computer scientists have been trying to create something that can converse with humans from as early as 1950s, however, conditions are now right for them to take center stage.
As a market chatbots were valued at $113 million in 2015 and is expected to reach approx. $1 billion by the end of 2024
What are chatbots
Chatbots are computer programs that can converse in the languages used by humans to converse with each other. To do so, computers need to understand what a human is saying and then respond with an appropriate text that a human can understand.
What is the current state of chatbots
Technically: Bots can be either rules based or based on learning (machine learning / deep learning) based. The rules based bots are designed to handle limited scenarios and they are programmed to specifically handle those scenarios. Learning based bots, understand and respond in natural language and learn from the interactions.
It is a lot easier to create a rules based bot and there are many online platforms that allow you to quickly create such a bot. You could see many such bots functional on Facebook. These platforms allow you to create from basic to very sophisticated bots. Though creating a customized bot that can understand the user's intent and respond with relevant content takes a lot of hard work and it requires a lot of data as well has competency in technologies such as Natural Language processing and Machine/Deep learning. For getting a taste of a what all can be done with a rules based chatbot try talking to our chatbot here
How they can be used by businesses
- 1. Personalization at scale:
A well implemented chatbot can revolutionize the ecommerce and other spaces. Think of a scenario where an online assistant is available to help you with all your queries and can personalize your experience on the website. However, poorly implemented bot can very easily ruin the experience as well. A lot of first stage chatbots are just converting the menu structure of the website into chat interface which obviously will not take your site too far. So a good suggestion would be start small, meaning only expose limited functionality on the bot but do it well and build on top of that.
2. Gain insights into customer behavior:
Every one gets used to old features on the websites and know how to avoid those traps. For example, users instinctively know how to cross out that email signup pop-ups ubiquitous all over the web. However, if a sales rep chatting with you asks you for email id you feel obliged to give that information. A properly implemented chatbot can collect a lot of useful user information.
3. Customer Service:
Bots are already being used in Customer service for sometime. Most of the companies have a lot of data around frequently faced issues by customers. They are using that data to train Machine leaning algos which are powering those bots.
4. Newer modes of engagement:
Chatbot provides a new way to connect with your customer, and as all web marketeers know and recommend, there is always a first movers advantage with things that come in new. So, embrace the new technologies and use it to your advantage.
5. Be present on all platforms where your users:
Chatbots allow you to be present on most of the platforms where your users are. And if implemented properly one central engine will power conversation on all platforms such as your website, social media such as Facebook, Twitter, Skype, Slack, Telegram etc. So, if implemented properly you could be connecting and conversing with your users at scale just about everywhere.
6. Learn as it grows:
A useful bot would be one that learns from its interactions with users. And learning process requires a lot of data, not only on how users interact with your bot, but also what all they want to know about your services. So, the sooner you start the better chance you have of getting ahead of your competition and based on data you will collect chances are you will stay ahead in the race for long time.
Bots at CodeFire
At CodeFire we are actively working on creating bots that can:
- 1. Making our own lives easy:
- a. Imagine a bot that can schedule all your appointments, sends emails to all participants and updates calendar based on natural language text such as "Schedule a Skype meeting with client X next Monday at 3:00 pm IST" and all this from your most used chat interface such as Skype.
- b. Imaging a bot that can quickly create tickets based on your discussion with clients from Skype chats and assign to correct developer
- c. Imagine saving the files at correct location so that the file is available when needed
- d. We are working on one such digital assistant, it in-fact already does 2 of the 3 use cases mentioned above perfectly!
2. Creating a deep learning and NLP powered platform that would allow you to converse with your customer on most of the platforms including your website.
Posted by admin on June 29, 2017