Write to us
Contributors
Writers
Designer
Artists
Concept and Direction
Enquiry
Talking to customers and understanding their woes has become more complex than ever. Thus, the solutions have become more technology-oriented, and new age tools such as Artificial Intelligence (AI) and machine learning have come into the picture, just to handle those special customer woes. In fact, customer engagement has also become trickier than ever. In this time of social media and technology-educated customers it has become more and more difficult to remain disengaged when the customer is looking for a solution or a new product.
MoEngage is an AI powered customer engagement platform for marketers and brands that help companies such as Samsung, Vodafone, Flipkart, Airtel, etc. to maintain customer engagement in such a way that they get more sales and more visits from buyers. This, according to Yashwanth Reddy, Vice President of Sales at MoEngage, is a win-win situation for both buyers and brands.
Interviewed by Avanish Tiwary
/ / How has customer engagement changed with respect to AI and the chatbots?
In terms of the overall template, it has become a lot more interactive than in earlier days. If you see how companies such as IBM work on customer engagement, you will notice that in the name of customer engagement, all they do is react. What I mean to say is that their actions lack any sort of proactive action or taking a lead, per se. Their idea of customer engagement is “if users come to our site, we will send a particular kind of email; if they spent a short duration of time we will send a different kind of email”. This is not action-oriented customer engagement; this is pure reaction.
A couple of years ago when we started we also used to do exactly that, but we have since completely changed the kind of engagement we do now. We thought it wasn’t the most effective way to do this. We wanted to be in a position where we were able to reach the right users using the right channel at the right time. That is the Holy Grail of marketing. And that is where machine learning and Artificial Intelligence (AI) have been helping us a lot. Over a period of time, we tried to understand the behaviour pattern of each customer and then tailor the communication according to what they like, how they want to communicate and then reach them on the channel they are comfortable with and regularly use.
But with our new product, we have figured out which is the right channel to which to send advertisements to particular users and the channel to which to send new product alerts, etc. We have even optimized and figured out the right time to send these emails and advertisements according to the customer’s lifestyle. Using machine learning and deducing theories from the data available, we know which particular customer is more likely to buy if we send emails at a specific time of the day. We help our customers increase their sales by implementing knowledge we get from machine learning and AI tools.
/ / Do these customers just spend more time on the portal or also buy more?
The way we process the data that we get from our AI tools, both conditions are achieved—people come to the website to spend time browsing products, and as they see more and more products their basket size also increases. We put pure science behind it.
As a company, I want my customer to spend more time on my website and I also want each transaction to be of higher value. Let’s say you have an iPhone. By seeing your past shopping data, shopping behaviour, etc., it’s really easy for me to up-sale an iPhone cover to you or an iPhone app, for that matter. The job of selling goods to customers by knowing what their past purchases have been is easy if the sale quantity is less. Say a person who has bought a kilo of Surf Excel will run out of Surf Excel within 20 days or so. It is very easy to sell to this person. But the brands we work with sell millions of products in a month. And that is where AI and machine learning come into the picture.
/ / Before AI and machine learning were the companies losing money?
Absolutely, right? I will give you an example. In the newspaper the toothpaste advertisement that I see and the advertisement that you see is same. Now the advertiser has put out an advertisement without knowing who is going to read this, what economic or social background the reader is from. At most the marketer has targeted the kind of newspaper the advertisement will be printed in; that is, if the advertisement will go to a Hindi or an English newspaper. But that is not enough to gauge if the money paid for the advertisement is useful and will be read by the target user.
What AI and machine learning do is see your buying behaviour and show the product related to it. If I know your last phone was in the range of Rs.5,000-10,000 and that you moved from a government sector job to a private sector, I will show you the phone of higher range. By not taking these things into account, companies lost a lot of opportunities. With the help of technology, now companies are able to use this to their and their customers’ benefit.
/ / Who benefits more from the AI and machine learning intelligence, customers or brands?
Both of them benefit from this. The reason is that as a brand I am able to make people buy more and more, drive more sales and at the same time reach out to my customers effectively. For users, since we have their purchase behaviour data and we make intelligent deductions with the use of machine learning, customers don’t feel spammed when we send them product infomercials or emails. The engagement has gotten better and they genuinely feel interested in the products we show them because we know they are already looking for something like that and that they can afford these products as they come under their purchasing power. So as a buyer, I don’t feel that I am being bombarded with useless product emails; rather I get more useful interactive experience.
/ / How large scale has been the usage of AI and machine learning by companies?
Well, this is at a very nascent stage of adoption in India. I would say it is being used most by enterprises only and they are experimenting with it. Vodafone, Indiabulls, Thomas Cook, etc. are using it with great success. I think we are still scraping the surface with the amount of data we have and it has a long way to go with more and more interesting use cases to come as we go.
/ / What is stopping the companies from using it?
Nothing is stopping companies from its usage. If you talk to any company now about the future of marketing, customer engagement and targeting clients, they will start the talk with digital transformation and the things that new age technology can do. So there is nothing stopping them as such. It is just that it’s a very initial stage and we have already, as an industry in the making, have started taking steps in the right direction. AI and machine learning have become a sort of revolution and buzzwords only in the last two years. Two years is nothing for it to properly evolve and get adopted at large scale for commercial usage.
From what we have seen by solving these customer engagement issues is that the enterprises have seen the results it can bring. Now they even have started to talk to their peers about it and a few of them are already in the process to start using it. So it’s just a matter of time before the usage permeates across the industry.
/ / How do you see the future usage of machine learning?
We feel that we have a lot of automation left to do. The way things are done right now there is a lot of human intervention in machine learning, and sometimes it brings hindrances to our work. More human intervention means more error and thus we sometimes miss our mark. So in the future, we would move towards making it totally machine-oriented. Over the next few years, we feel there will be a lot more automation and that will directly turn into high return on investment for companies and better suggestions from the machine and AI technology.
Just like when you go to a physical store human eyes can see what they want to buy. That is also going to happen to customers when they go to a website. For each customer, product placements on websites will be different pertaining to their buying preferences and will also be sorted by their purchasing power. In a way we are moving towards personalized websites for every customer.