Kannan department store has five branches in Coimbatore. Step into any one of them and the experience is eerily similar: crowded queues of people waiting at checkout, tightly packed aisles jammed with shoppers and aggressive price discounts. It can be stifling. There is no air-conditioning or mall-music here, but you’ll mostly find what you are looking for.

This is just one of the hundred places that people in the town go to shop. Elsewhere, in the claustrophobic commercial nerves of Coimbatore, wholesale traders, in dingy shops squeze next to each other like Papadums. Thousands of small, around the corner shops, stock up from here. Large open tent markets do brisk business selling vegetables and fruits. People don’t mind the mud and dirt to get to the shop.

Welcome to commerce in India’s small town. It is thriving, intense and fragmented. The opportunity it presents to an online retailer is large and growing, as discretionary income rises. There’s more: The per capita GDP in many Tier-II cities is set to rise three-fold in the next decade.

Unmet needs are accumulating.

But it is not going to be easy pickings for online retailers. The reasons are two:

1. Middle India is not homogeneous. It’s a loose collection of thousands of smaller markets each with its own unique culture, language, needs, motivation and interests. Good luck selling to a local shop owner with a website in English.
2. Middle India is homogeneous in its limitations. Middle India has poor infrastructure, has lower income and needs to touch and feel products. Getting e-commerce solutions to go above and beyond (winning the trust of a naturally skeptical consumer) under these constraints is a challenge.

Middle India’s heterogeneity demands micro-localisation from e-commerce and this could be inefficient. You don’t have the luxury of serving large monolithic consumer segments here.

Assembling locally relevant selection

A handful of brands determine consumption in certain categories. Be it Samsung in mobile phones or Zara in fashion, the needs of today’s Indian e-commerce consumer – primarily, the creamy layer – is easier to define, assemble and sell.

The story of small town India’s needs is very different. Even something as innocuous as the good old Lungi needs to be different for different people. “We dove into the markets and realised that the pattern and colour of a Lungi that sells south of Madurai is very different from a Lungi that sells in Kozhikode,” says Ganesh Balakrishnan, AVP Seller Services at ShopClues.

Customers in smaller towns are very specific about what they want. Product selection gets more unstructured, local brands begin to emerge (Kitex Lungi anyone?), preferences become unique and popular price points get extremely narrow (and often lower).

ShopClues, which differentiated itself from the rest of the e-commerce market by selling to small-town India, perhaps understands this better than anyone else. Five years later, having built themselves in specific markets in the north, they are looking south, hoping to break in – a journey that Ganesh is helping lead for them.

From unique kitchen utensils, to bed linen to fashion preferences driven by regional TV shows, Shopclues operates in the zone of narrow specificity for each market in addition to the larger selection. It’s a tightrope walk.

Amit Bagaria, VP of Business at PayTM, believes that the true value of marketplaces is yet to be exploited in India. “In order to go deeper into the country beyond the top consumers, large, open marketplaces are the only reasonable way to offer the diverse local selection,” Bagaria told me.

His words are a reflection of the ethos of PayTM which heavily draws inspiration from China’s gargantuan, multi-layered commerce ecosystems. TaoBao, its largest marketplace, supports 7-million merchants on its platform (by comparison Amazon India announced that they’d crossed 1 lakh sellers on the platform a few months back).

Marketplaces haven’t quite turned out to be the business model manna from China.

Marketplaces haven’t quite turned out to be the business model manna from China.

Flipkart’s much feted move to become a marketplace in 2013 was quietly whispered into oblivion in just over a year. Theoretically, they still are a marketplace. But they pick and choose their sellers and still have a lot of inventory led sales. Amazon arguably put more ‘market’ in its market-place but it’s an elaborate simulation of the inventory heavy, close-to-retail approach through premium sellers and its minority-owned Cloudtail.

Snapdeal continues to remain the poster boy of marketplaces but loses relevance every passing day. As FactorDaily wrote in August Snapdeal’s market share has come down from 24% in October 2015 to 14% in the quarter ended June 2016 in a market which ships about 26 million units every quarter. Perhaps Snapdeal’s hypothesis of price over trust and convenience is not entirely true for India?

Bagaria is unfazed when you rally evidence against the success of true marketplaces. He says, “Inventory led models work better for the top 10% consumers.They buy structured products, demand fast deliveries and great deals”

But for small town, middle India, you need to have a thriving marketplace. Stock up on the wrong type of Saree at your own peril.

In 2012, Shopclues was uncool and it continued to be for the next several years – for the elite, urban consumer, that is. Early on, the platform decided that they’d play true to the unstructured, fragmented festivity that was small town India. Its managed marketplace experience was true to its name. It meant staying away from fund-sucking, glitzy mobile phone exclusives or big-sale blowouts.

Today, Shopclues is a new Indian unicorn that has built a small-town empire targeting $3 bn GMV by 2016. Here’s the clincher: It’s targeting profitability by early 2017. To be sure, Shopclues has been drumming up profitability since 2015.

It got here by taking the hard path of building supply and demand market by market: Physically exploring individual markets to identify purchase patterns, in-demand local products and preferences and replicating this online by bringing in sellers both local and otherwise (who may offer the local selection demand).

Today, as the shiny layer of the market hits its limits, the Shopclues’ model may be coming of age, accumulating merchants, winning markets and quietly showing the way that a managed marketplace, built ground up may be the only true way to tame middle-India.

That also makes Shopclues a prime target for acquisition by a larger, well funded ecommerce player.

Supporting small, local sellers

Here’s a confession from one of the leading marketplaces in the country: “Greater than 90% of the sellers on our marketplace platform don’t make a sale.”

There is enough evidence to believe this is true for nearly all marketplace platforms. The story of e-commerce marketplaces providing an avenue for lakhs of sellers is largely the job of spin-doctors who feed the media frenzy around e-commerce.

Yet, it would be impossible to build local selection without a backbone of small sellers. But getting them to sell online is tough. Here’s why:

1. Small and medium sellers have limited working capital. Short term credit is either unavailable or is extremely difficult to get.
2. The burden that large e-commerce platforms often place on sellers including the commissions, relatively long payment cycles (7 – 14 days) as well additional costs that of long lead-time of returns (up to 30 days) make it unsustainable.
3. Most sellers are hardly found on online retail platforms. For instance, more than half the sellers on most platforms don’t see any customer traffic.
4. Sellers don’t realise the impact, aren’t comfortable using the tools at their disposal and sometimes do not have the resources to build their presence online

Most seller platforms aren’t built for small sellers but for the few, well-funded large sellers who accumulate inventory and offer them to customers. Now let us take these problems one by one.

First, the working capital credit problem.

Non-Banking Financial Companies (NBFCs) have been a boon to sellers, opening up faster, easier access to short term loans. Nearly all e-commerce platforms have leveraged them for seller credit (CapitalFloat and LendingKart offer loans as low as Rs 1 lakh to sellers).

But, we aren’t even scratching the surface as far as addressing small seller needs.

The needs of a small seller are hardly addressed.

Those who are eligible on the platforms need pedigree (running business on platform for a year and in some cases even 3 years in the case of eBay) and sufficient captive-ness on the platform (high ratings and reviews) in order to be eligible for support.

Remember those 90% sellers not making a sale? If we do the math, we end up with a really small set of sellers who are even eligible for support. Even for these sellers, despite the promises of rapid disbursal and quick approvals, the process remains cumbersome.

“No one has truly leveraged data fully yet to offer fast, low-friction instant loans for SMEs,” says Madhu Sudhan, co-founder of Loanzen, which is focusing on enabling invoice-based discounting loans to sellers in hours by getting rid of all manual, paper-based processes and verification. Loanzen has achieved sub one-hour loan disbursal cycle using data (from aggregators, public sources and mobile usage) to arrive at predefined credit limits disbursed through a mobile app led process.

These possibilities have come up because of Aadhar-based verification, better integration with data and more intelligent programmatic decision making. This is an area e-commerce firms need to adopt with greater intensity rather than look at it as an additional service for the sellers. The ability of small sellers to have access to capital instantly for short term working capital needs could create a huge impact on the ecosystem.

All of this brings us to point 2 – platforms squeezing sellers. To be fair, let’s blame the business model.

Long payment cycles and escrows for safety (in case of customer issues, returns, etc.) often make it unviable for sellers (many small sellers covering specific markets or products) from a capital liquidity point of view if not the eventual margins.

Could a share of all the discounting investments be directed toward offering more attractive terms to sellers? Perhaps, there is a greater need for integrating seller data across platforms and even payment providers? This could give proactive view of seller credibility which could in turn be used to make sellers more liquid in the platform.

Yet again, we look to China. Ant Financials, China’s largest fintech company, shows the way here by offering collateral-less, instant loans (in a few seconds) for small sellers based on data from its payment service Alipay (claims to have 450 million users today).

All of this has to work in line with educating the sellers.

Bagaria of Paytm feels that local sellers would especially need dedicated support to get them going, “It is expensive but will pay off in the longer run,” he said.

Language localising access for customers & merchants

Imagine this: You come across a strange device lying by the side of the road. It is filled with colourful boxes and shapes all written in an arcane medieval language you don’t understand. Pushing them creates more visuals and sounds. Now, go shop.

The experience is not very different for many consumers for whom the smartphone device offers a strange new interface experience in an alien language – English.

Only a third of the country’s population can understand the language. Enabling local language powered e-commerce (from awareness to product discovery to purchase to support) in essential if it has to open up to middle-India.

The 45 million Indians who browse in local languages today do so despite the limited local language content. The need is strong.

Yet, language localisation in e-commerce has been a non-starter, mostly. Snapdeal, one of the early movers in this space, is yet to see a massive traction for its language websites. Even Shopclues, which differentiates itself with a primarily focus on small-town India, is only now on the verge of launching a Hindi language version of its site. Flipkart and Amazon have largely stayed away from any deep attempts at language localisation, so far.

Arvind Pani, CEO of Reverie Language Technologies, which provides localisation solutions, believes that the inflection point for language localization is 12-18 months away. According to him, organisations are largely playing a waiting game.

Besides, it is a complex puzzle to solve for e-commerce firms. With crores of product listings (that tend to grow everyday), the dynamism of the various discovery features and intelligent algorithms that work off them, localisation needs to be a real-time, intelligent feature.

“What businesses fail to see is that localisation is already happening all around them. Open up any app marketplace and look for top entertainment / news apps in India and you are likely to find a large proliferation of local language apps,” says Arvind Pani.

Today, content firms like Facebook and Google have begun pursuing language localisation aggressively (Google’s latest announcement that Google Assistant will be available in Hindi being one of the many threads). Most media providers and platforms have aggressively localised. Perhaps, the inflection point may be here sooner?

The benefits of language localisation are more immediately visible on the merchant-side, reducing friction and in turn improving customer experience. And parts of it are easier to execute for the e-commerce firms.

Balakrishnan of Shopclues has witnessed this first hand. When the company made its training videos in local languages, their seller engagement shot up rapidly.

That’s content – the easier piece of the puzzle. The bigger benefits are likely to accrue when platforms that the merchants use, from logistics to warehousing to inventory to payment, offer local language interface.

But language localisation cannot be seen as a growth hack. The support for the consumer and merchants has to go beyond just the e-commerce platforms – to the devices that consumers use, payment interfaces and supporting infrastructure.

Rakesh Deshmukh, co-founder and CEO of Indus OS, says that the company has built its operating system to support language content even on third party apps. “For example, a consumer can use our features like Text to Speech and translation/transliteration on most applications on the phone,” said Deshmukh.

These fundamental localisation of access stack may indeed help open up consumption and later commerce for the small-town india.

Retailing in local cultural context

Merely having a local language stack (and localised interface) isn’t enough. Ultimately, retail and consumption are closely tied to local needs and habits and this is where e-commerce firms have a long way to go.

Local language media still accounts for nearly 90% of media consumption in middle India. Any awareness building initiative has to extensively exploit these channels in order to succeed rather than large national campaigns that are monolithic in their messaging, offers and highlighted selection.

Here, merely dubbing a national advertising campaign into local language isn’t likely to work. The concerns are likely to be different, so are the types of products and periods of peak sales.

Seasonality and cultural context also play a huge role in driving offers and promotions. Local festivals and events are significant drivers of commerce and this is unique to different parts of the country. For instance, retail seasonal spike (could be as high as 30% of annual sales) in Kerala happens during Onam (several weeks ahead of Diwali which is the predominant retail season up north).

Yet, e-commerce firms today aren’t making enough of an effort as they chase instant scale and rapid topline growth.

Personalisation

Personalisation of platforms will be crucial in order to break through the clutter.

While platforms today support rudimentary pin-code based availability of products, a highly personalised, local and relevant customer experience is still a non-starter. Today, it is largely limited to offering fine tuned recommendations (even this needs to go a long way). But beyond that, the platforms remains largely static.

There is data becoming available for ultra-personalisation.

India’s consumer behaviour is on the verge of becoming data-rich, especially as the digital payments layer begins to take effect. “Payment play should be more correctly called a data play,” feels Jitendra Gupta, Founder of CitrusPay. The payment technology company was recently acquired by Naspers’ PayU for $130 mm in an all-cash acquisition. He is betting on the fact that the rich customer data that payment providers have will be unmatched and extremely valuable, especially for middle India whose digital footprint has been small so far.

Ultimately, localising the experience (from language to selection to promotions) for every consumer on top of the other localisation stack will truly cater to the small-town consumer and open up new frontiers for Indian e-commerce.

---