Hyperscaling a digital business in India : Part 1

Building a digital business in a market where revenue earned per transaction shrinks as your offering democratizes

This essay is related to the broader theme of finding a template for successful digitization in India that I have been exploring here. In this part, I will talk about how even cases of successful digitization of economic transactions in India are full of firms that haven't been able to hit profitability due to the unique challenges of scaling their business in India.

If you are a venture investor putting extremely high risk capital in Indian start-ups, I would assume that you have a definite growth path charted out in the industry that your invested company is trying to disrupt. You would expect the first 12-18 months roughly to be the era of tech enthusiasts and early adopters, using your product and providing valuable feedback for better future iterations. At some point after that the demographic or policy tailwinds that you would have bet on would likely kick in, driving the mass adoption of the product sold by your portfolio company. Your product democratizes, deepens its base and adds new features to keep its customers loyal.

Eventually the idea is that the average revenue per new user will increase, as your product becomes more valuable to each new customer because it already has incorporated their needs through feedback from older customers. In short, your average transaction size per user should go up in the long-run, as you become more embedded and more valuable to their lives.

Of course, I am assuming here that you are a rational venture investor.?

How does the evidence from the three most digitized economic domains of our lives stack up against this theory??

The chart below shows the Average Transaction Size for UPI and NETC-FASTAG since December 2016 in nominal terms. I have indexed both prices at 100 for December 2016.?

Increasing democratization, increasing sachetization of transactions

The numbers fall off even more steeply when we adjust them for inflation. The chart below measures the former two ATS in inflation adjusted terms. I have also added data from SEBI on Average Trade? Sizes in the F&O and Cash equity segments, adjusted for inflation. As before, I have indexed all these numbers to 100 for December 2016.?

For me, the more surprising aspect of this analysis was? the ATS for Cash Equity and FnO segments. Usually, as markets go up, given the preference for investments assets remains the same, you would expect some increase in average trade sizes, as people would want to pay more to acquire those shares/ futures baskets in order to take bigger bets on the market. That has not happened.

Even as prices are up 40% in the past seven years, the average size for an equity trade has gone up from INR 22K to INR 26K. That of an average FnO bet has gone up from INR 663K to INR 757K. Clearly, both have lagged inflation.?

In the electronic toll collection space, the average size of a transaction per tag has fallen from INR 4221 in December 2016 to INR 708 in October 2023. Contrary to theory, the new customer is also not someone who is somehow passing toll plazas more frequently. Average number of transactions per tag is 4. It was 15 seven years ago.?

In the next part of this essay, I come up with some theories that could possibly explain this common trend of falling ATS, its uniqueness and what it implies for firms operating in these vanguard digital sectors.