E-commerce industry in India is booming by leaps and bounds. The early movers in e-commerce are expanding their existing portfolio, as well as entering into new ones, whereas newer players are coming up with niche offerings. All these players are struggling with a lot of challenges to make their position in the online market space. To know more about these challenges, InformationWeek spoke to Ashish Shah, COO & Founder, Pepperfry.com.
The launch of Pepperfry was in response to an unmet customer need for a wide range of quality furniture and home products in India at one place. In the last 3 years, Pepperfry has shipped products to customers in more than a 1000 Indian cities. This reflects the growing trend of how Indian customers are readily adopting the online medium.
The key is delivering a smooth experience despite the geographical and infrastructural constraints existing in India. When Pepperfry opened for Business in January 2012 we made significant investments towards building strong logistic networks.
One of the biggest challenges is the standardization of the customer experience. Our investments have been towards controlling all aspects of customer experience starting from pre-sale merchant listings to post-sale logistics and customer service. This helps us provide a standardized and superior buying experience while retaining our marketplace model.
Stringent quality check measures enable us to ensure high quality standards. Almost every item sold on Pepperfry first arrives at one of our logistics hubs where it is checked for quality, before being packed and shipped.
In case of made-to-order furniture, our team monitors and tests products at each stage of preparation by manufacturers; this ensures that only the highest quality products reach our customers. Deliveries take place through our own delivery network or through experienced 3rd party logistics partners. In the rare instances where a customer is dissatisfied with the product or the product is damaged in transit, we stand by our customer promise and either offer to fix the product (especially in the case of furniture) at the customer premises itself or refund the purchase amount to the customer.
The biggest challenge is poor infrastructure and supply chain. E-commerce as an industry can grow exponentially provided the infrastructure and the ability to scale up and deliver products to every corner of the country develops at the same rate.
Poor ‘last mile connectivity’ due to missing links in supply chain infrastructure is limiting the access to areas where a significant portion of online customers reside. For a country as big as ours the Internet penetration has covered only 10% of our population. This is in stark contrast to the global average of 35%, and much below the average of the developed nations at 78%. We have a population base, which is big enough for a thriving e-commerce industry but the e-commerce potential at the moment is limited by the broadband penetration.
The plethora of local tax regimes are also a huge challenge for e-commerce. The implementation of GST is now a pressing need in the seamless world that technology operates in and any budget measure which indicates an early implementation of GST regime would be a value addition to the business.
For a specialized vertical like furniture large item distribution poses as a huge challenge. There is no prototype in the online industry for a vertical like furniture that can be imitated. It was a huge challenge that we overcame by pioneering our own “Large Item Distribution model” which hasn’t been attempted by any other company in the past. By making significant investments in that area we have built a hub and spoke model that covers a majority of our orders. This comes from our business orientation that “when you take risks, you will make mistakes. But you should be the best at fixing them.”
Analytics and data form an integral part of every crucial decision. We collect data through a few well-established third-party tools and from internal reports, which we have built-up over a period of time. Data helps us analyze trends and decide what is working for the business. It enables us to eliminate subjectivity from the decision making process and thus, take clear quantifiable measures and calls.
At Pepperfry, we track every possible data from marketing campaigns, visits, user actions, events, funnels, etc. A/B testing helps identify the preferences of our consumers and gauge their thinking process. As time progresses and technology enables us to gather and process more and more information from our users, in light of readily available data, we believe that any other alternative for decision making will cease to exist.
We have recently started working on Apache Kafka + Storm deployment. At Pepperfry, we realized that we have a large amount of data being consumed as metrics besides event notifiers. We decided to go in for a sort of a scalable and distributed data bus to ensure that the messaging layer does not collapse, when gigabytes of data are being passed through the system.
We felt that we needed help in distributing messages under heavy load, with low latency and a real time processing layer, which can help classify the piece of data/message for further processing.
Apache Kafka is a well-known cluster based distributed queuing system that offers strong durability and fault-tolerance capabilities. Apache Storm on the other hand, is a distributed real-time computation system. It reads the data off Kafka and process the data stream as programmed, by processing the data into information, which is then to the various systems for storage and querying.
We have already deployed an analytical engine based on Elastic Search, which drives certain slice and dice kind of information analysis required by our various teams. Certain aspects of the website tend to become monolithic when we start combining inter-dependent features into one module. As Pepperfry has grown in terms of business volumes, we felt the need to scale in terms of h/w as well as features vis-à-vis maintaining low costs, which led us to take the route of micro-services.
It enables us to reduce code inter-dependency between new features as well as gives us the opportunity to test certain aspects of the website before assimilating into a module.
At Pepperfry we have a very strong technical team that works on developing technology in-house. Most of our technology requirements are fulfilled by that team. They have built a very unique code deployment tool along with an automated server scaling layer which depends on a whole lot of metric info gathered from the live environment, to auto-create and deploy online.