Digital winner-take-all platforms: Hype or reality?

Over the past four years, entrepreneurship in India has seen a focus on copying Silicon Valley digital models. The chief features of this start-up landscape included high capital consumption by a mere handful of players to artificially hasten industry consolidation. Other dimensions include questionable investor returns and limited job creation. Going forward, there is a clear need to consider more capital-efficient and effective approaches to foster millions of start-ups, rather than merely a few thousands.

Why has so much attention and investment focussed on digital start-ups? That investors merely ‘followed the herd’ is the most commonly discussed behavioural explanation. However, given that billions of dollars have been invested across multiple rounds by several investors, a widely held economic assumption may underlie the big bets. The oft-repeated mantra of ‘winner-take-all online platforms’ in investors’ public statements and popular discourse represents the key underlying economic rationale. Winner-take-all outcomes occur when the dominant intermediator earns higher incremental revenues per user and the lion’s share of available value. This monopoly payoff is very different from that of a competitive oligopoly.

The Economics of Winner-take-all:

Winner-take-all online platforms may emerge through a combination of scale and network effects. More users in a network can increase the value for other users, for example with facsimile machines, telephones, gaming consoles, alumni groups, etc. These are the networks effects.

It is also true that digital networks, relative to physical networks, can enjoy more rapid scale-up and geographic reach at substantially lower capital intensity, as well as negligible marginal costs to serve incremental customers. These are the scale effects. When added up, these effects can exhibit the relatively rare yet powerful phenomenon of increasing incremental returns on investment.

Rather than assuming that winner-take-all outcomes are inevitable in all online platforms, it is important to critically examine their relevance and applicability in specific contexts. For example, in the Indian context, will e-commerce and online cab hailing indeed end up with winner-take-all platform outcomes? Are the actions of investors in shaping such outcomes in the long-term interests of the economy and consumers? Or is this more a case of a few smart investors deliberately funding cash-burns of billions of dollars in operating losses to artificially raise barriers to entry and induce rapid consolidation to cosy duopolies?

The latter appears a plausible explanation of the fairly rapid observed consolidation of online cab-hailing aggregators in most markets. For example, consider Uber-Lyft in the U.S., Uber-Grab in Southeast Asia and Uber-Ola in India. Only a handful of big pockets—Tiger Global, Softbank, Ali Baba and, Tencent—have funded the billion-dollar plus annual cash burns of each player.

From an economics perspective, market structure and particularly monopoly outcomes depend on the fundamental advantages of scale economies, network effects, control over key resources, customer stickiness and/or regulatory barriers. Are any of these fundamental advantages upheld in the specific context of Indian e-commerce and online cab hailing?

Google and Facebook appear to be winner-take-all platforms in online search and social media respectively. However, they undoubtedly compete directly in the digital advertising market, which after all, is the important money side of their platforms.

The argument here, however, goes beyond the mere semantics of winner-take-all versus ‘winner-take-most’. Unlike the cases of online search for Google or social media forFacebook where ordinary consumers do not pay anything for the service, in e-commerce and cab hailing, consumers pay for each transaction. In such a pay-for-service models, will consumers and regulators tolerate higherduopolistic prices? Will consumers not seek price comparing substitutes?

Despite the recent consolidation in the Indian context, how difficult will it be for other new entrants to enter into the online cab hailing or e-commerce market? Two scenarios are possible. In the first, these new entrants will also have the appetite to fund large operating losses while chasing the assumed future pot of last man standing valuation gold. Alternatively, they could enter once prices are high enough to make operating profits sufficiently attractive. For example, concerted duopolistic behaviour by organised retailers Coles and Woolworths in the Australian supermarket segment eventually prompted additional entrants and increased competition (Rice and Martin 2017). Similarly, in the Indian context, players such as Paytm or Jio could be seen as future new entrants into the cab-hailing segment.

If millions of people search on Google for the same keyword at the same time, the search volume certainly does not negatively hinder users in any way. Indeed, it might actually help users, perhaps with better machine learning effects owing to more data. The same holds true for Facebook.

In contrast, beyond a threshold level of users, as Eisenmann and co-authors argued in a 2006 Harvard Business Review article, there can be negative “same side network” effects in online cab hailing. Drivers would not like having too many other drivers at a given location node at a given time. Similarly, riders may not like having all the riders in the city using their platform as that too can increase the probability of delays or surge price hikes at choke points. So in the case of online cab hailing, the prospect of perceived monopoly power and negative same-side network effects beyond a point could hinder the move towards a single dominant platform. Further, several recent analyses (most prominently, Horan 2016) suggest that Uber has no economies of scale, so their attempt to fund large operating losses to drive out competition has little fundamental economic rationale.

For e-commerce, the argument is that strong customer service along with the natural scale and scope monopoly effects that are necessary to make last-mile parcel delivery to homes profitable will eventually force a single dominant player. This argument seems farfetched. In most markets, FedEx, UPS, DHL and the nationalised postal services continue to slug it out even today. Given the challenges of managing large armies for local delivery, there is also the logic of outsourcing to multiple local aggregators of last-mile delivery agents. In other words, there may eventually be only a few national/ global players, but the prospects of significant local competition remain.

Elsewhere, online platforms such as Twitter and LinkedIn appear to have achieved winner-take-all outcomes in their niche. However, this has not translated to significant monetisation or valuation benefits, compared to the scale of U.S. and Chinese tech titans such as Google, Amazon, Facebook, Baidu, Alibaba, and Tencent. Will investors value true user engagement or merely continue to chase growth in user numbers?

In the early stages of e-commerce and online cab hailing, the heavy discounting and subsidies have actually helped enhance consumer welfare. However in these business models, consumers pay for every transaction. Any sustained attempts at duopolistic/monopolistic price increases will provide competition regulators a clear indication that their intervention is required.

Could e-commerce and taxi-cab hailing in India reach stalemate outcomes, where no player can earn returns above cost of capital? Will the large Indian digital players reconsider their prevailing focus on burning cash to acquire millions more customers through excessive discounting? Could the focus shift instead to superior consumer engagement with higher value customers? These are interesting strategic alternatives, worthy of deeper reflection.

Tejpavan Gandhok is clinical assistant professor of strategy and entrepreneurship at ISB.

With permission from ISBInsight, ISB’s flagship research periodical. For the complete article go to: isbinsight.isb.edu