ChatBot Assitance to Businesses

Chat Bots- An Over Hype, Yet A Necessity

The direct co-relation between AI (Artificial Intelligence) and over-hype seems to have existed for over a decade now. Complete general AI, where a machine could respond to customers with human like interactions is a dream that is not yet achievable, despite the market promises of such a possibility.

Why Is Complete AI Bot Not Possible?

Chat bots are built based on contexts which help create sophisticated voice interaction scenarios. Context help the bots differentiate between vague phrases or phrases that could have different meaning. For example, if you are listening to a Song from, lets say One Direction; and you tell your bot "I want to listen to more of them". Now the bot needs to be smart enough to understand what "Them" refers to. A Smart bot could either play more One direction songs or songs which similar listeners like too. However, understanding the context is not always that easy for even humans, let alone bots mastering it anytime soon.

Subject context (remembering what the conversation is all about or tracking the conversation history to know what the users are talking about) or environmental context (like the location you are in) are relatively easy for the bots to handle. But external contexts (Remembering events like Trump winning the election, Brexit, or a celebrity dying) could be a little tricky as the bot would then need to also know the interests of the user well to narrow down on the context. And the most difficult for Bots to handle would be the social context- knowing the emotional state of the user (might not be in the mood for humor and might not be the right time to joke around) or the level of friendliness/ formality to express and understanding Idioms and slag is a different challenge all together and it would take quite some time before Bots reach that stage. Last stage would be what I like to call Honesty. Will bots be able to utilize the concept of what we call "White lies" is something we need to wait and watch.

Then Why Do We Need Bots?

Brands like Sephora do a great job on the social media (Sephora has 2.4 million consumer interactions across facebook last year). With all consumer brands focusing on social media as the key platform for marketing and customer service; keeping up with such large global communities is a mammoth task for the brands. The Pareto principle holds true even here; 80% of the queries are repetitive and can be answered by a bot with artificial intelligence. The 5 key areas where bots can play a crucial role for the brands:

• Customer Service: When a customer needs an answer immediately, and that question is not easily discoverable, a conversational UI can offer the most direct path to satisfying that need. Customer service is a great example of this use case
• User Training: For brands like Gillette Venus for example, Customers (especially Teens) can learn about Shaving and get some tips instantly using the bots. Even Tutorial Videos and links to Website pages can be pushed to the user targeted specifically to their needs.
• Purchase Support: Bots can be used to help customers with product recommendations and simple purchase tasks like adding products to basket and check out can be handled by the bots. This brings Conversational commerce to the forefront.
• Repeat Purchases & Re-targeting: Bots can send information directly to a messaging app, continue where the last conversation left off, and send well-timed personalized messages at whatever time interval you want.
• Consumer Analytics: Bots can be used to analyze and categorize the comments in real time, understand consumers better to personalize messages and even flag abusive content in real time.

Fun Fact: During the 2016 US election run-up, the use of Twitter Bots make up an estimated one in five election related tweets.