What are chat bots? Why are they such a big opportunity? Why the big companies like Facebook, Google and Microsoft are into it?
This blog will answers to all these questions, plus overview on how to develop a chat bot for your business – so continue reading…
What are chat bots?
First thing first – let’s understand what are chat bots?
Chat bots are the services which are powered by rule engine & sometimes artificial intelligence which can be interacted via chat interface. These interactions can be straightforward, like asking a bot to give you a weather report, or more complex, like having one troubleshoot a problem with your internet service.
On technology front chat bots needs rules engine, big collection of data, artificial intelligence like natural language processing & of course the support of all the companies who has chat app (Facebook messenger, kik, WeChat, Skype ) etc.
Why are chat bots trending?
Business need customer & these days’ customers are more on mobiles. The best way for a customer to interact business is by mobile app. But a smartphone owner installs zero app per month – they only use 5-10 app on regular basis. Also building a good app is hard, get it to the hand of the user is must harder. There are 3 million of apps on both the major app stores – it’s overwhelming. People are tired of trying out latest-greatest app.
So what business has to do if it wants to interact with customer? – This paved into “The Rise of chat bot” . One of the big attractions of messenger’s chat bots is that customer need not install another app nor go to the website and login. They can just send a message to chat bot and it would reply back in the same chat window with relevant information.
This has become a new trend for the customer front companies in industry like e-commerce, travel & tourism, media and news etc. These bots reads the customer message, process it using artificial intelligence and return information with links to the product / news pages.
Big companies are interested in chat bots.
Facebook owns 2 messaging app – Facebook messenger & WhatsApp. Few signals say that company is increasingly looking for ways to make money from Messenger. Even though Facebook have not completely ruled out advertising as a source of revenue, the more immediate revenue source can through charging business such as hotels & travel companies for providing reservations & other service through the chat app.
Another messaging service, kik which is quite popular have introduced “bot shop”, where users can ask question to bot about weather, help in online shopping etc. Slack a messaging service used by business has recently partnered with Taco Bell to introduce a Taco Bot. This bought helps the users to order from local outlets. One in five WeChat users have added credit card & bank account information so that the person can check their balance, pay bills, send money to friends etc.
There is a big opportunity for all companies who own a messengers platform to act as a personalized ordering system or information request system. In other words, they can bring in orders or requests in a big numbers.
Your new personal assistant is with you 24 x 7.
On the other side, we have Apple Siri, Microsoft Cortana which are different form of bots. Behind the scene, the technology remain the same – artificial intelligence, rule engine, big data etc. Facebook is also testing with its own digital assistance called “M” that can answer question, perform task in ordering food etc. It uses a combination of artificial intelligence and input from human overseers.
Bonus: How can we develop a chat bot?
Chat bot can be created in any language – PHP, .Net, Python etc. On 100 feet view it
- Use the API from messenger apps to send & receive messages
- A user input processor
- A rule engine and/or artificial intelligence engine
- A output processor
The overall process is simple. Develop can get the in inputs of user using API provided by the messengers. The input processor will convert the message in required format by rules engine or artificial intelligence system. The AI system and rule engine will produce the required results based on the user inputs. This is the hardest part to implement. Next these results send to output processors to format the message as per the defined rules of messenger being used.