 "Srikanth Velamakanni")
One fine Sunday morning, 29-year-old Nishant Malhotra woke up to non-stop message beeps from his smartphone. Irritated, he picked up his phone. The first SMS read, "Stop everything and start shopping: Get extra 40 per cent off only today; 900-plus products. Code WOW 40." While this offer came from Gurgaon-based online shopping portal Jabong, the next message he checked had information on a similar promotional offer running on fashion portal Myntra. There were myriad SMSes from a handful of other shopping portals urging him to make the most of his day-off from office.
Malhotra, an avid online shopper, began his Sunday early but without a whimper of protest.
A s the next phase of consolidation kicks in with Flipkart's acquisition of Myntra, e-commerce companies realise that the one metric that will help them achieve growth is focusing on revenue per visit or RPV, which is a combination of the conversion rate on the website and the average order value. Mind you, this is one step ahead of the single-minded focus on conversion rate visible at the time when the e-commerce industry was still young. The smart ones know that while they could probably double their conversion rate by cutting all prices in half, this would likely have a negative impact on the bottomline by reducing the overall revenue generated. Watching revenue per visit ensures that the host site is increasing the rate of conversion without compromising revenue.
Leading e-commerce companies like Flipkart, Myntra, Jabong and FabFurnish already claim they are shipping around two to three and in some cases up to five products for each order received on their website. The whole idea is based on simple math really. Experts peg the cost of acquiring a customer by e-commerce companies has come down - at Rs 300, down from Rs 1,000 two years back - while the cost of servicing an order has gone up with an increase in manpower and infrastructure costs. Says Ravi Vora, senior vice-president, marketing, Flipkart, "There is enough noise about online shopping today and the best part is that each player is not required to put efforts separately to draw attention. Today more than 60 per cent of shoppers on Flipkart are repeat customers."
In such a situation what will separate the men from the boys will be the way a player clubs his offerings and services the last mile to make the whole process more efficient. Praveen Bhadada, senior director at Zinnov, believes, "Analytics will decide the prospects of growth for e-commerce players." In other words, how well a firm knows who its customers are, what they want and how to urge them to spend more every time they walk into an online store is key. And what will arm online store managers with this knowledge is customer data harnessed from the website and other sources and crunched and put in a shape that helps in customising each offering.
Discounts are juicier than ever
Coming back to the point on driving volumes by offering discounts, many players have a similar discount strategy: lure customers into spending a fixed amount of money by dangling the juicy bone of high discounts, sometimes up to 50 per cent. Now there are two ways of offering discounts. One is the old school general sale in which every consumer gets a fixed discount at a certain point in a given category. The other, and the more new way, is about creating different catalogues for different customers. To put it simply, a catalogue sale is an occasional deal available on a given assortment for a stipulated time period. For example on June 23 both Myntra and Jabong ran a discount deal of 40 per cent on a set of products compiled in one catalogue. Says Praveen Shah, co-founder and managing director, "When we give incentives to shoppers, the chances of repeat purchase go up significantly."
FabFurnish follows a variant of the ticket-size principle to design discounts and drive volumes. The company claims that currently the average number of items on a shopping basket is two/three. The baskets are divided as 'furniture' and 'non-furniture'. While the average order size of a furniture basket is around Rs 10,000, non-furniture basket, which may include bed and bath, decor, lighting, kids and baby products etc - stands at Rs 3,500. In the last two years, FabFurnish has tweaked its discount strategy completely. If earlier it was offering discounts based on the ticket size -that is, the higher the ticket the bigger the discount - now it has fashioned lucrative offers on smaller ticket prices as well.
Alongside, it has chased this set of buyers relentlessly by improving its product recommendations. The principle of recommendation works like this: Apart from suggesting brands and offers, the site will also prompt other categories of products that a buyer could buy along with the original product on the list to avail of an extra discount. The results, the site claims, are as expected. About 30 per cent of the shoppers clicked on the recommendations and conversion rate went up by 20 per cent. Says FabFurnish co-founder Vikram Chopra, "If one does not put some constraint on the order value, the revenue will go down. Also, it is the best way of increasing the number of items per basket."
To drive volumes and cross-category impulse purchase, Myntra has been running what it calls 'basket promotions' for a year now. Says the company's COO Ganesh Subramanian, "Picture a scenario in which a consumer has come on the website to buy two products. After making the selection her order value comes to Rs X. By adding one more product of lesser value, she will be able to claim a Y per cent discount on her order. In most of the cases we have observed the consumer ends up buying the third item." What he means is that in doing so the consumer usually experiments with a new category. Myntra claims the number of items per order has gone up by a count of three products in the last one year. Catalogues created for women have driven volumes for the company. Similarly Jabong has seen a big jump in sales by cross-selling accessories.
Getting the logistics right
As leading e-commerce companies exit the phase of customer acquisition to take on the challenge of customer retention, logistics will be crucial. Says Shah of Jabong, "When per-order value goes up along with the number of items, it is viable for us to pass on the savings to the customers." This is achieved by driving efficiencies in logistics.
Let us try and understand the math. The cost of delivering two items of the same size to the shopper's doorstep will not be radically different from what it takes to deliver one unit. In this scenario, if the ticket size on a given order goes up by, say, Rs 1,000, an e-commerce company can log savings of up to 20 per cent on its delivery cost, say experts. How? Take just one element: call centre charges. When an e-commerce company outsources call management, it has to pay a certain amount. If the number of calls remain the same but the order value associated with a call goes up, it means same workload - and therefore the same fee - for the call centre but higher realisation for the e-shop.
While most online shopping companies that started off with an inventory-led business model have cut down heavily on stocking inventory and moved towards the managed marketplace model, order aggregation is forcing them to re-evaluate their strategies. Take this example. Suppose a customer in Chandigarh has demanded two products from a website that works on the managed marketplace model. If it has to source these two items from two different merchants located in, say, Surat and Delhi, it is unlikely that both the items will reach the customer on the same date. Add the shipping cost the e-commerce company incurs. Where does it leave loyalty and efficiency?
Subramanian of Myntra - the portal which hopes to be profitable by 2015 - says, the website tries to forecast as best as possible but yes, it doesn't get it right 100 per cent of the times. "There is a gap between demand and forecast," he adds. "But our split order percentage is in low double digits."
But there's a catch. Delivering more items per order will demand more investments from logistics partners. Says Sanjiv Kathuria, co-founder and CEO at e-retail delivery fulfilment company Dotzot, a DTDC company, "If the weight per shipment is 1 kg or more, we will have to look for a transport solution other than bikes." To leverage the network of DTDC and accomplish timely deliveries, Dotzot is planning to bring some of the best global practices in logistics to India. 'Click and collect' is one. As part of this, online shoppers will be able to pick up and return their orders at multiple booths set up by various players.
Aggregating orders is one answer. But the task is easier and faster for companies that stock a major portion of the inventory. With a number of promotions lined up during any given week Jabong has managed to increase the number of items per order by 25 per cent from last year. To service its biggest market of Delhi NCR faster, the company has opened four packaging centres in the NCR itself. In this way it is able to deliver within 20-24 hours of receiving an order.
In all this shopping portals are following in the footsteps of their brick and mortar predecessors. Devangshu Dutta, chief executive officer of specialist consulting firm Third Eyesight, sums up the trend succinctly: "E-commerce companies in India have to focus on the principle of low price and low cost. Global players like Amazon and Walmart have grown by offering lowest prices and keeping their operational costs low. Promotions drive repeat purchase that eventually make up for lost margins and this is no different for e-commerce companies."
| Expert take |
|
|
CUSTOMERS DON'T LIKE TO BE SOLD BUT LOVE BUYING AND WELCOME ANY HELP THEY CAN GET. Good sales people know this and focus on fulfilling customer's shopping mission by offering relevant recommendations. In the online world, analytics aims to substitute this responsive salesperson with three important differences:
- We can leverage a "perfect" memory and "infinite" computing power to find what is relevant to the customer at the moment
- We can reconfigure the store layout and "show" customer only these relevant things
- We can leverage reviews and buying behaviour of friends/peers
With every action customers are conveying information about their attitudes, preferences, life-stage and socioeconomic status. If you buy a Prada handbag, it doesn't take genius to know that you are an upscale fashion conscious consumer who doesn't mind splurging. If you regularly buy a full basket of groceries, but never buy meat, odds are higher that you might be vegetarian. What if this kind of customer understanding can be developed algorithmically at internet scale--over millions of customers, billions of transactions and interactions--we can create rich customer understanding across several dimensions. Companies have developed solutions to understand customers, decode their "genome" and use it to make better recommendations.
These solutions leverage statistical or machine learning techniques. Collaborative filtering, a machine learning technique, became famous when, in 1998, Amazon acquired Junglee Corp for a staggering $183 million to strengthen this capability.
Once you understand your shopper, you must also learn what she is trying to do. What search terms did she use? What is the time of the day, day of the week and her geo-location? When you search "red roses" instead of "cheap flowers", Google results and Ads are very different. In the first instance, Google shows fewer ads than in the second instance where Google "knows" you are more likely to buy. Understanding customer context can dramatically improve the relevance and efficacy of product recommendations.
Once we "know" the shopper and the context, we can hyper-personalise her experience of the store with relevant products and offers that meet her needs. This is a problem of plenty – analytical techniques discussed above can prune and prioritise these offers.
When you see a store having sections like "people who bought this also bought this" or "people who considered this product ultimately bought this" or "you browsed this product, you might also like to consider these products", the recommendation engine is at work to personalise the page. The higher the relevance of recommendations, the bigger is the size of the shopping basket.
We are social creatures and follow what our friends/peers do. If relevant information about our friends is presented, it can increase conversion and help shoppers take the leap of faith required before high involvement purchases. Advice from other users is usually seen as credible, unbiased information and improves customer conversion and size of the customer's shopping basket.
If you have wondered why online stores want you to login with your Facebook id or provide them with your social media permissions, it is to enable them to influence you with what your friends bought or liked.
Online stores frequently have limited time offers (for instance, additional 15 per cent off for three days only) and basket size based offers (for instance, additional 10 per cent off if you spend more than Rs 5,000) to create time pressure and improve size of the customers shopping basket. Customers can get hooked to these tactics and stores might find it difficult to wean customers away from these expensive tactics.
Stores can experiment by understanding customer price elasticity and offering every customer a unique price at which they are willing to buy. So, in theory, a store can charge one customer a high price if she is price insensitive and offer another customer enough discount to make her buy. When such pricing is permitted by law, it can still lead to customer angst and loss of trust.
Customers are smart and have a choice to take their business elsewhere. If an online store can understand them and their context deeply, hyper-personalise their shopping experience and earn their trust, they can hope to inspire customer loyalty.
Srikanth Velamakanni
co-founder & CEO, Fractal Analytics
co-founder & CEO, Fractal Analytics
