4 questions to ask before designing a “Chatbot” for your business

The number of smartphone users in the world is forecasted to grow from 2.1 billion in 2016 to 2.5 billion in 2019. As our devices start getting smarter (and humans dumber, pun unintended) it would be fascinating to see where the next wave of technology will take us on how the human race will use internet as a whole. As the number of smarter people (I mean smartphone carrying people) increased, the number of interactions they do with various brands has increased drastically and more and more companies are using chatbots to ensure constant availability for their customers.

Though some feel that the term chatbot has recently entered the public lexicon, they have a much longer history than you may expect. It all began in 1950 when Alan Turing, an English computer scientist, published an article titled “Computer Machinery and Intelligence”. He outlined the Turing test, a way to measure if one was speaking to a human or to a chatbot. This in many ways is considered to be the beginning of AI. Starting from there, now voice search, voice assistants, and even augmented reality have been integrated into our everyday use.

How are chatbots being used by Marketers?

Chatbots are now being used heavily for customer service and online ordering and we as customers are increasingly getting used to minimal human intervention with the various brands we deal with. To share a recent personal experience,

I ordered food online
After the food was delivered and the delivery boy had left, I realised that one item was missing
I complained on the app by following a few easy steps and choosing the correct dropdown options
The missing item was delivered to me within a stipulated time

In all the above interactions with the app, apart from the delivery boy, no human interaction was involved.

Looking at the future, as more and more data is generated; the chatbots will keep becoming more and more intelligent. Head of Evangelism for Bing, Christi Olson says,

“Gartner predicts that by 2020 people will have more conversations with chatbots than their spouse”, she adds that “the chatbots of the future don’t just respond to questions. They talk. They think. They draw insights from knowledge graphs. They forge emotional relationships with customers”

A few years ago, this statement would have belonged to a Sci-fi movie, but is completely believable and achievable now. The only point of contention here is about having more chat with a bot than with your spouse. Some may find it concerning, and some may find it to be a relief but that I leave on you to decide.

How one should use a chatbot for their company?

Like all marketing strategies, a brand needs a clear plan before investing in any customer interaction platform. So before jumping on to invest into a chatbot, ask a few questions to yourself:

1.Do you have enough data/consumer insights to build a meaningful chatbot?

When a consumer lands on you site/app and has some queries or steps to follow, he should have meaningful options to choose from. If you introduce a chatbot without an exhaustive (semi-exhaustive) list of data point at the back end, the consumer would probably have a bad experience with the chatbot and leave with a worse impression about your brand than what he came in with. In short: Just having a chatbot on your site will not make your brand look smarter if there is no meaningful information/interaction that it is providing. So first step would be, to build an exhaustive set of back end data points collected from user interactions, for building the chatbot.

2.How complex is the service/product that you are delivering?

If it is a simple service/product which is being delivered every time, for example, an app for ordering three types of special diet breakfast cereals which would be delivered to the consumer. Carefully identify the limited number of failure points in the service and the backend would not be too exhaustive to start with. So don’t delay a chatbot introduction, introduce with the optimal number of failure points that need to be taken care of. As you would now understand, for more complex products/services, all failure points need to be identified to build a meaningful chatbot. In this case, a better approach would be to create a bot just for a few initial steps and give manual intervention for complex issues. The customer service executive builds a repository at his end which later help in building a more exhaustive bot.

3.What is the business outcome you want to drive?

For instance, does the bot enable you to reach untapped markets and create leads? Or, it could be about customer retention and customer service, if you want to provide technical support without having to wait for a live technician to be available. You don’t have to create a list of extensive business outcomes, just the key ones (even one key business outcome is good too). One you have a clear objective in mind, the design of the full user flow towards your objective will be clear in your design.

4.Which are the steps that will removed/made easier by introducing a chatbot

One of my friends loves to plan trips for a group of people. He will look for places to travel, check the budget, make the travel arrangements etc. However, he hates the additional back and forth communication to make sure the schedule of the trip matches with all members of the group. So for him, rather than making it easier to plan the trip, if someone designs a bot to match schedules of all the members, is a much higher value add. So, identify the pain points of your consumer and try to make them easier. Don’t try to replace the tasks that the customer loves to do in the first place.
Hope you enjoyed the views shared. Happy chatting.
PS: In my last blog, had mentioned that Zappos is reported to have a 75% repeat customer base.
You can read the blog here
If we look at the shoe buying habits of US consumers, as per this 2015 report, they buy 7.5 pairs of shoes per capita per year. With this information, Zappos will know which of their consumers are buying all their shoes from Zappos while which of them (although repeat) just buy two out of their total purchase of 7.5 shoes in the year.
Thus, they can separate the loyal base also into Evangelists, Strong Loyalists, Loyalists and so on and can create a different approach within this repeat user base to deliver the relevant WOW.
Authored by Siddhartha Banerjee, Co-founder,GoCrackIt