Paul Writer launches The Rise Of The Modern Marketer,
Video interview series of CXO’s highlighting best practices and trends in the use of technology for marketing.
Co-Founder & CMO
Use of analytics in your marketing decisions
There are two to three places that we use analytics. One is the way we use custom acquisition. We try to figure out which of the channels and media give us the most number of new customers. For example, if you’re using Facebook, Google Scmsu or digital ad campaigns, we would track conversions by channel. We would look at which channel the customers are coming from, what kind of price are we paying for each new customer? And what is the eventual retention of each customer that is coming from a particular channel. This data would really allow us to figure out which of the channels is the best channel for us and therefore, optimise our marketing mix by channel. We do the same thing even for Atl. For example, one of the things that we did recently was to use data coming from our T.V advertisements. So we know what time our advertisements are getting released and on which channel. We actually track the customer visits to the site and app as well as the number of conversions at that time. We took an interval of about ±3 seconds after an ad was released so as to check what happened to the customer visits and conversions. We could correlate which channels and which insertions were working well for us. Over a period of time, we know which channel is giving us better conversions and the time of day that works well for us. As a result we’ve been able to harmonise our entire marketing campaign even on T.V, where we have been able to reduce costs by about 30-40% and still get the same amount of visits and conversions.
“We’ve integrated our entire social media channels, email marketing channels, all our onsite clickstream and all the transaction history of a customer into a single view. This data helps us trace customer patterns and make intelligent marketing decisions based on that.”
The other place where we use analytics very strongly is in terms of life cycle management to customers. We can look at who our loyal and new customers are and how they behave. We also look at the in between customers who are trending to become loyal and find out their behaviour. What we do with this, is we learn about the kind of items and auto values they buy, how often they visit and what pages they visit. We use this data to make sure that our offerings to them is customised. For example, one of the things we realised is that for trial customers showing them offers works best because they get more confidence and get to know that they’re getting products at a great price. For all new customers, the site would actually show more offers than it would for the regular customers. For the regular customer, the site is customised to show more categories. Therefore, they start buying a wider basket. The idea is to really tune your offering based on what stage the customer is in his/her lifecycle with Big Basket.
Big Basket is about convenience so how are you using data in giving better customer experience?
One of the biggest USPs that the online grocery store offers is convenience. Can I get my order delivered at home? I don’t have to spend time and I don’t have to go to the store and spend hours in the aisle. One of the things we’re trying to do is to make sure that this trend transcends to shopping online. One thing that we’re trying to create is the fastest possible shopping experience on site. Our goal is to get regular customers to check out in two-three minutes. Today, they take 18 to 20 minutes. How do we do that? We’re building what we call a ‘Smart Basket.’ A Smart Basket is a device or a programme which is for any existing customer. After a customer has placed about 10 orders or so, it will start recommending what the customer should order. It will look at the past order history of the customer, predict what the customer needs at that point in time and actually create a basket of what the customer should buy. It does this by looking at standard deviation, interval of ordering, what categories that they buy and it also makes intelligent recommendations. It does not necessarily assume that you are going to buy what you have before. But it actually calculates, for example, you’ve bought rice three days back and your normal ordering frequency for ordering rice is once in a month. It is not going to recommend rice to you. It’ll recommend a product that you may buy at an interval of three days. Therefore, the Smart Basket is intelligent enough to calculate and compute a customer’s basket every time a customer logs in. It is based on the past history of the customer, how similar sets of customers buy and based on what the customer is potentially likely to buy. Therefore, the idea is really that if the customer uses a Smart Basket, they should be able to pick 80 to 90 percent of the items they regularly buy in just a couple of minutes. This makes it extremely convenient for customers and we believe that if customers use this, they will get hooked onto online grocery buying. It will make it a lot simpler and easier for them.
The other place that we really use analytics is to figure out what UI and Ux to present to a customer. We look at the data of how customers navigate the site, use the app and figure out how much time they are spending there. Are they finding any difficulty in finding something on the site? Are they getting enough information? Are they spending a lot of time finding a particular set of category of products? We designed the site around what we think is the best way a customer navigates. For example, one of the things we discovered when we were doing this is that new customers don’t know what kind of guarantees we offer. So, one of the first things that customers see now is that we offer no questions asked return guarantee. This reassures new customers. The idea is to really bring messages and navigation which is customised to a particular set of customers. For this we use analytics and data to look at how the customers behave to deliver this.
Biggest challenge with respect to Marketing Data and Analytics
I think there are two big challenges. One is the sheer volume of it. A lot of times, marketers get lost in the details because you’re generating data if you’re an online business on every transaction and a clickstream every time a customer visits. So you know how the customer is navigating the site, how much time he spends, what items he adds to the basket and what he deletes. Whether he applies a coupon or not, what does he do after applying a coupon and what payment instrument that he uses is known. So the amount the data that you collect is voluminous. Figuring out, therefore as a marketer what your priorities are, what data you want to use and how to use it intelligently is itself a key task. One of the things we’ve done is that we’ve really identified what we think our customer facing priorities are. The top 3 things that we think are important to a customer are:
- How do we expand? How do we make sure that the customer has a great experience on site?
- How do we expand the range of offering that the customer buys from?
- How do we ensure that the customer buys from almost all the categories that we sell?
We are using these priorities to really define what data we want to work with and how do we present and use that data? That is one challenge, the sheer volume of the data. The other challenge comes with analytics today and online businesses that customers interact with you in multiple formats. They may tweet about you, post on Facebook, send an email, call your customer service, they will navigate on the site, do an order or even give feedback to the delivery guy. Trying to get a 360 degree view of a customer is very difficult. How do you build this 360 degree view, how do we know and integrate all the feedback channels of communication that a user typically uses? In our case, we’ve integrated our entire social media channels, email marketing channels, all our onsite clickstream and all the transaction history of a customer into a single view. This data helps us trace customer patterns and make intelligent marketing decisions based on that.
Listen and manage customer feedback on social channels
We’ve done two or three things to make sure that we manage social media well because it is a extremely important channel. In fact, increasingly social media is becoming the customer service channel for most companies. And customers have taken to tweeting and use Facebook or Instagram to post pictures of the food that they receive. Therefore, they have become very important customer service channels. One of the first things that we did was that we built a dedicated team to address this. So we have a team that addresses all social media responses across all the different channels and they handle only that. They have expertise in that, they know that they have to respond in a particular tone of voice and they know the kind of response they’re supposed to give.
Secondly, we’ve APIs integrated into all the channels. Regardless of whether it is the Google Play store, Twitter or the Facebook page; we get all the responses into one single customer service channel. Hence, the customer service reps are able to see it and respond. Thus, this team is able to respond in real time.
Thirdly, we’ve set strong KPIs, say a Facebook post needs to be responded in three minutes, a tweet needs to be answered in five minutes or a comment on the play store has to be answered in thirty minutes. We monitor how many posts are answered within that time, how many posts were repeated and how many interactions did it take to resolve a particular complaint. Ideally, the idea is to you should resolve everything in one interaction. So if you’re having multiple interactions, it means that the customer service is not working well and people are not able to respond well. We use that experience to go into our database to figure out how to respond better and provide better customer service.
Therefore, we are integrating all the customer service channels into one single view. And we use an integrated team to address them in time.
We’ve also trying to make sure that instead of going to a social media channel, they can just tweet from the app if they have a problem. They would then receive an instantaneous response.
Future of online grocery business in India
The online grocery business is fast becoming a mainstream retail channel. I think, in the next 4/5 years, you’ll find at least two or three large multibillion dollar companies in the country. I’m pretty sure that big basket will be one of them. I also believe that the largest online grocery companies will be as large as the physical retailers that we have today. This is happening rapidly because for one, customers are adopting it as a mainstream channel. We are already seeing it as our market shares are as high as physical grocery retailers in the top metros. I think that trend will take an even stronger foot over the next few years.