Concerns about startups that yearn for network effects

Business Standard, 2 May 2016

Everyone loves the computer revolution. We extol the greatness of the startups and the shiny baubles which we get to play with. There is a dark side to this, however. A lot of the big financial success in technology startups is about `creating network effects'. Each such success story has an element of an impregnable monopoly that yields supernormal profits. This raises concerns about competition.

There was a time when computer companies were workmanlike creations. Infosys or TCS are great software services companies. They were built through slow hard work over decades. In the case of TCS, most of the founding team was absent when the most remarkable financial performance was obtained. This kind of slow hard work is unfashionable in the world of technology startups.

There is ample fuel in the computer revolution to do many useful things. It is possible to improve products and processes in almost all parts of the economy using the riches of cheap CPUs, fast networks, decent batteries, an array of sensors, all coordinated by Internet protocols and implemented using various flavours of Unix. This kind of work is slow and careful: it involves fully understanding an old economy firm or industry, and injecting high technology into it. Done well, this kind of work can generate good returns. It cannot generate magical outcomes.

Technology startups have no patience for this slow path. They dream of getting up to a valuation of $100 million in three years. How can this be done? Sometimes, this can be done through radical scientific innovation. This is relatively rare. The elixir that every technology startup is instead looking for is `network effects'.

A business has network effects when the presence of each additional user improves the quality of the product offering. As an example, when one more person uses Google mail, this makes Google mail stronger, as it becomes easier for Google to block unwanted or dangerous emails. That most people use Google mail has, in and of itself, created a strong entry barrier which holds off rivals like Apple mail. Once these network effects set in, product attributes matter less. Google can, in principle, get away with an inferior Gmail for quite a while, because it's protected by the network effect.

It's useful to distinguish between indirect and direct network effects. In a product like Google Mail, the functionality resides in the software, and the functionality is enhanced by the addition of one more user. Here, the network effect is indirect. But in something like Ebay, which is a retail market, the functionality is the user. In and of itself, Ebay is useless to the first user. What's useful to the end user are the wares being put up for sale by others, or the orders being placed by other users. Here, network effects are direct: every additional user directly enhances Ebay. Indeed, there is no value in Ebay without the buy or sell orders being placed by users.

The `marketplaces' of E-commerce parlance are `exchanges' in financial economics. When you go to NSE, the liquidity that you obtain is directly made by the others who have placed orders at NSE. Every order that you place at NSE directly makes up NSE's offering, which is the high liquidity of the order book. Without orders from users, there is no liquidity at NSE. Access to other users is the NSE offering. Once liquidity comes about at an existing exchange, it is very hard to compete with it. There are only a handful of examples, in world history, where a new exchange could take on an incumbent, if we exclude the US where peculiar financial regulations forced a fragmented market.

In the E-commerce world, the name of the game is to find an opportunity to setup a network effect, and ride it to glorious financial returns for the founders. From an economics point of view, there is something uncomfortable about the spectacle of one after another firm that is so keen to achieve a little impregnable monopoly, after which the normal rigours of competition do not apply.

Capitalism works because firms in competitive markets have to ceaselessly push the boulder up the hill. A constant stream of innovations are required for enhancing the product and cutting costs. The best that can be obtained by a good management team is reasonable returns on equity. This essence of capitalism, this Calvinist ethos, is lost once a monopoly is setup. Shareholders and managers then enjoy supernormal financial returns, and freedom from the pressure to constantly improve the product and reduce costs.

In network industries, the profit motive encourages managers and investors to do many perverse things. As an example, speed is more important than product quality. The firm with 2000 users beats the one with 200 users. A weak product that's released early wins when compared with a great product that's released late: there is a bias in favour of speed and not product quality. Another odd thing is the role of capital. Money has become the raw material. The firm that subsidises users on a bigger scale gets more users, and then the network effects take off. Competitive advantage is then about obtaining more and cheaper capital from the global financial system, and not building a better product. Put these together, and the sloppy product that's backed by the big money wins.

People who work in the field of technology startups find this depiction of a subset of the startup space (the quest for impregnable entry barriers through network effects) to be familiar. This very quest for monopoly places these startups in the gunsights of public policy economists, whose dharma lies in identifying market failures and addressing them. In the public imagination, the supernormal returns of technology startups are always born of amazing engineering innovation. I worry that too often, what's going on is more mundane: technology firms are obtaining supernormal returns by building little monopolies.

The computer revolution offers great opportunities to improve products and processes in all industries. Too many technologists and investors are, however, too high to engage in this slow hard work. The game has become one of finding the quickest path to network effects, after which you have a monopoly. The reward for this is supernormal financial returns and laziness. The spectacle of young men and women, who are imbued with a get-rich-quick ethos, and dreams of laziness, is disturbing. Economists need to reinvent competition policy so as to address the market failure (market power) in this landscape.

Back up to Ajay Shah's 2016 media page
Back up to Ajay Shah's home page