Some thoughts on ONDC

The Ken has a long read on the Open Network Digital Commerce (ONDC) initiative of the Government of India which is an open (anyone can join) and inter-operable meta-market for buyers and sellers and third-party Apps. 

The comparator cited for ONDC is Unified Payments Interface (UPI) which since it launch in 2016 has rapidly become the host for two-thirds of India's online merchant payments amounting to six billion monthly transactions. The ONDC's Chief Business Officer Shireesh Joshi is typically bullish, 

“Double e-commerce value in five years, a 100-fold increase in the number of retailers online, and triple the number of people who buy online.” To put that in numbers, 30 million sellers and 300 million shoppers by 2024.

ONDC's supporters think they have taken care of the potential scaling challenges,

Bringing sellers onboard is a lot more effort-intensive. “Buyers know how to buy online. These are sellers who have never sold online. They need to know how to digitise catalogues, maintain inventory online, and use different packaging for every order. They need a lot of training.” And that is the job of the seller aggregators like GoFrugal and eSamudaay... what about customer experience and trust in these platforms’ service standardisation? 

Anup Pai, co-founder of eSamudaay, says these are solved problems in commerce. “Eventually, the net value here is in the movement of food from a local restaurant to a local home. You just need to have somebody who knows the restaurant and its location. So, I’m purely looking at it from a financial perspective. Profitability bana sakthe hain isme, but pura cholesterol mein jaa raha hai (you can make profits in this, but right now it is consumed by cholesterol—overheads),” he says. His solution is to create “10,000 Swiggies” in every district... Pai says a decentralised local digital marketplace model has the scope to create multiple companies valued at $1 million and some $100 million ones across cities...

The way Pai describes eSamudaay is not very different from Shopify, the online marketplace that connects retailers directly to customers. He calls his company “Shopify in a box” since he wants to replicate a Shopify-like marketplace in every district. However, unlike millions of sellers in India, Shopify sellers are digitally savvy. They log into Shopify and can have a digital shopfront within minutes. They don’t need assistance to come online. “In India (especially tier-II cities and beyond), you can’t do digitalisation without a digitising agent. It needs an agent. So, you need to incentivise the agent to be a partner in the seller’s journey,” he says. Pai wants to turn these agents into entrepreneurs who would run the local digital marketplaces. “There are plenty of 45-50-year-olds who have been in business, have connections, and still have the energy to build on the ground. Their aspirations are to take care of their family and not build a unicorn. We’re providing them templated business solutions.”

... 24-year-old Srilakshmi Acharya, who runs the Udupi operations for eSamudaay... the way a Swiggy would, but with eight people—for merchant onboarding, customer support, digital marketing, accounts, operations, and delivery... eSamudaay charges a 15% commission to fulfil an order compared to 25-32% of Swiggy and Zomato... For updating inventory, eSamudaay charges an onboarding fee of Rs 1,500 ($19). It does the hard job of mapping a store’s inventory, after which the store can update it.

What they overlook is the sustainability and unit economics involved in all these interfaces. What if the binding constraint for small sellers is not digital literacy and training, but unit costs? Would the mom-and-pop retailers be willing to pay the price required for eSamudaay to operate a sustainable business? What's required to establish credibility and trust, and change behaviours among these mom-and-pop retailers? Can eSamudaay figure out a scalable model to establish trust and onboard retailers in an ultra-low cost manner, something semi- and unskilled labour market matching platforms have not managed to do for nearly two decades? 

Are small Swiggies commercially viable entities? Can small Swiggies be created as a templatable business solution? If we are looking at creating 10,000 Swiggies, are there that many entrepreneurs and investors (actually we need ten times that many to start with) spread out across the country with the risk appetite, entrepreneurial nous, and diligence required for the purpose?

Sample this snippet about what goes on in the mind of a mom-and-pop retailer

Ajesh Pai (who runs the small shop) says paying a 1% commission to eSamudaay will be worth it if it can bring him more customers. “We don’t even give a free bag. Margins are very important for us.”

One of the comments to the article hits the nail on its head

The comparison with UPI can only go so far. UPI is simple. It has only one SKU – money; only the “order value” and the leading +/- sign change for every transaction. And the settlement rails were prebuilt. Obviously, ONDC is far far more complex. I’m not clear who ONDC’s primary customer is and what problem they solve. The current set of digitally savvy buyers is well served.

The driving force behind technology start-ups is growth. It's believed growth creates the conditions for sustainable business development.

1. At the intensive margin, growth helps acquire customers and transactions and maximise network effects. Apart from adding to the top-line, it also boosts margins in a technology business.  

2. At the extensive margin, growth creates unforeseen emerging business opportunities. An Amazon of 1995 or even 2005 could not have foreseen some its business lines of today. This is especially so in platform businesses. 

There are some assumptions behind the growth push

1. The higher margin arising from the regulatory arbitrage between the regulated brick-and-mortar business and its largely unregulated internet variant.

2. Once customers are acquired and the business grows into a behemoth, it would wield significant enough market power to be able to raise prices and increase margins.

Both these assumptions are now being tested. The former due to the aggressive anti-trust push back, and the latter due to the drying up of cheap capital and the increasing focus on profitability. 

A few observations in this context:

1. Foremost, the comparison with UPI is misleading given the vastly different nature of the two activities. Consequently the ambitious scaling aspirations of ONDC betrays an ignorance about the size and nature of the India's e-commerce consumers, and underestimates the challenge of establishing credibility and onboarding sellers.

2. The vast majority of retail transactions in India are done by the poor and lower income population. They're extremely price sensitive and the convenience of e-commerce may not be enough to get them to spend a little more on e-commerce transaction fees. The opportunity cost of time for the vast majority of Indians is not high enough for them to avoid shopping physically. 

However, this is not to gloss over the fact that despite being a small rich and middle class as a share of population, they are a large enough market in absolute numbers to support big e-commerce players. But these e-commerce players too should be cognisant of the size of the market and build their business models accordingly.

3. Building businesses involving innovations is not about growth. Instead, it's about getting the offering (product or service) right and figuring out a sustainable business model through a process of experimentation and iterative adaptation. It's about optimising costs and getting the unit economics right. 

4. Finally, and this I think is an important but hardly discussed aspect, an ONDC-like platform could generate two sets of dynamics. One, it eases market frictions, lowers costs, and expands the market. Two, it allows price-sensitive buyers to shop for the cheapest product and cheapest among sellers. 

The latter poses problem. In an already price-sensitive market which is primed to minimise costs and margins, do we need more margin reduction and race to the bottom? We already have examples from the telecoms and airlines market where such race to the bottom has taken the market to the precipice. I have blogged here about how Make for India conflicts with Make in India for the World.

None of these observations should be seen as an argument against ONDC. Indeed ONDC is a great step in the right direction. But its growth and expansion should be more thoughtfully planned with greater understanding of the complex nature of the market and the economic activity, and the broader dynamics of market competition. Most importantly it should eschew the ideological approach followed by growth-at-any-cost startups. 

Its objective should be to create a robust public good and not mimic some canonical startup growth model. The success of ONDC should not be seen in terms of how many tens of millions of retailers and hundreds of millions of customers use it. Instead it should be about how much it would contribute to productivity increase and expansion of the aggregate production possibility frontier by formalising and enhancing the incomes of producers and traders in the informal sector. 

Something like Alibaba's rural Taobao, which would demand entrepreneurship which builds on the platform. Do we have such entrepreneurs?

Update 1 (18.12.2023)

The Ken has an article on the problems with ONDC Apps, highlighting the troubles of Namma Yatri, a ride hailing App in Bangalore developed by Juspay Technologies.