The side effects of network effects
I first heard about network effects from Venky Harinarayan in 2002. I was in my early 20s and worked for one of his portfolio companies in Mountain View, California. My ignorance was written on my face when he first mentioned network effects in the context of peer-to-peer technologies. He explained it to me like only Venky can: “If you were the first user of Napster, it would be useless for you. Can’t share or receive anything. Then one day your friend wants to share a music album with you on Napster. They tell you how you can get music from millions of strangers. You are almost forced to join. That force is network effects.” Since then, the idea never left me.
I understood how hard it was to get network effects going while working for a company making mobile instant messaging and voice-over-IP apps well before iOS and Android were born. Then I spent a decade watching multiple flavours of two-sided marketplaces from the inside—first connecting thousands of local small businesses to lakhs of consumers in India (Chaupaati), then running a business connecting tens of thousands of product suppliers to millions of online buyers in India (FutureBazaar), then working for a business connecting tens of thousands of restaurants with tens of millions of diners in the US (OpenTable), and then watching the world’s leading travel company connecting millions of accommodations with hundreds of millions of travelers worldwide (Booking.com’s parent Priceline Group). Two-sided marketplaces have strong network effects as well. For example, if OpenTable only had a few restaurants to book a table at, it would hardly be useful. Since OpenTable has exclusive inventory of most of the popular restaurants in the city, you have little choice but to use it to book a table. Similarly, top restaurants have little choice but to list on OpenTable since that is where diners go to book a table at their favourite restaurants. That’s network effects.
Fast forward to 2015. I met three entrepreneurs in their early 20s, each celebrated for being hyper-funded in quick succession by a top-tier VC and then a highly coveted global fund. I quizzed them separately about the endgame of their hyper-growth with deep negative gross margins per transaction. After their back was against the wall in the argument, each of them dismissed my scepticism as ignorance of their secret enigmatic weapon—network effects. Had I failed to understand the network effects in two-sided marketplaces after having experienced them at scale for a decade or were they hanging on to the thread of hope while being as ignorant about it as I was when Venky first told me about it? I thought long and hard. This is where I landed.
Indeed, each of these businesses has network effects at work. Indeed, each of them is experiencing a force that is drawing consumers and suppliers into an irreversible habit. Unfortunately, in the most undesirable ways. The same force that they hope will spiral their companies upward with a virtuous cycle is spiraling them downward viciously.
Consumers have been paid to buy from them long enough that they stop transacting unless they are getting paid to do it. They have got into the habit of being paid through discounts, quick gratification and free conveniences. They have learnt to hack the system by getting discounts and cashbacks on the first and last transaction, and every other transaction in between.
Suppliers have flocked to join the network because they are being paid to sell through them. They flock to competitors at the drop of a hat and revert to old behaviour when incentives are withdrawn. Suppliers have got into the habit of being paid minimum guarantees, tiered bonuses and free technologies. They have devised ways to double dip into incentives, disintermediate the service after connecting with the customer (sometimes in the same transaction) and collude with friends to invent fake transactions out of thin air to make real margins.
Suppliers need consumers to unlock their incentives, and consumers need suppliers to unlock theirs. The consumer-and-supplier habit that has been created is a habit of being paid to transact on the network a.k.a perverse incentives. When perverse incentives stop, so do the transactions. A popular hyper-funded company that recently shut operations saw an 80% drop in transaction volume when they withdrew perverse incentives. In a country with experts in the sport of who-blinks-first, the start-up and its investors consistently hand over victory to consumers and suppliers.
It takes two to tango. With three or more, all hell breaks lose. Perverse incentives were an outcome of aggressive competition between multiple participants backed by a bottomless pit of funds rewarding topline metrics of demand, supply and transactions. Whenever a point of friction came up on either side of the two-sided marketplace, start-ups threw money at the problem and resisted adding controls that would throttle growth. If the growth stopped, funding would stop too.
These start-ups created a more convenient way for consumers to do things like shopping, getting a cab, doing groceries, booking affordable accommodation, ordering food, etc. They could not charge for the additional cost of convenience because customers would not pay. They had to eat up the logistics and inventory costs of returns because suppliers would not pay. Additional costs of this new way over the old way were absorbed by the start-up, with neither side of the marketplace willing to pay for it. Growth continued because they found investors who were willing to pay. Investors were willing to pay because they experienced success doing this in the US and China. Now they are learning that it is not the same.
In the US, consumers pay for convenience. Distribution of products across categories is efficient and homogenous. Suppliers use technology systems rather than human systems. This results in centralization of competitive advantage and few winners take up larger market shares. In China, manufacturing of products across categories is as cheap and scalable as it gets. Suppliers control product margins because they are producers, not traders. There is a nationwide infrastructure for transport and distribution, and policy framework that works like clockwork. The closed nature of the economy ensures that once the government and the few Internet conglomerates have made their bet, the ecosystem must obey their verdict. Expecting similar behaviour in a fragmented, local, unorganized, human-systems driven, open economy such as India is like yanking the thermostat of an air conditioner outdoors and waiting for the temperature to get regulated.
We are seeing network effects at play in a marketplace where both sides of the network have two characteristics that together form a lethal combination. One, high tolerance for pain. Asked to choose between pain and paying up, most would prefer pain and defer paying. Two, a highly evolved ability to hack the system when a participant in an established market category is willing to lose money. What appears to hyper-funded start-ups as the easiest way to own a large market in record time turns out to be the easiest way to burn a large amount of money in record time.
When consumers and suppliers in a two-sided marketplace are brought into the network through value, network effects will amplify the gains. When bought with money, network effects will amplify the losses. The force that giveth also taketh away. Network effect is a great power that needs to be used with great responsibility.
Kashyap Deorah is the author of The Golden Tap: The Inside Story of Hyper-Funded Indian Startups. He is an entrepreneur and investor who shuttles between India and Silicon Valley.
His Twitter handle is @righthalf