Platform businesses: Competing with Big Tech

When you're lost in the cycle of product development, marketing and customer support it's sometimes hard to see the big picture of the forces reshaping the structure of the telecoms industry. In particular, telcos are in an unfolding position of co-opetition with what you might call 'Big Tech' -- the IT technology, commerce and services giants. These increasingly overlap with telco functions. Many of these companies have platform business models. These create value for end users as well as upstream suppliers, and extracting revenue from joining the two sides together. Think Google, Amazon, Sun Microsystems or Salesforce.com. Companies like IBM specialise in construction and servicing of platforms, even if they don't always feel the need to own them.

We strongly believe that telcos need to form a platform around their own unique assets. But what drives the economics of platforms, how should the telco platform be positioned against those IT platforms, and what lessons can telcos learn from them?

Mass-produced IT processes for a mass-production world

It's often been suggested that various so-called network industries exhibit increasing returns to scale; whether or not you accept Metcalfe's law, it's empirically obvious that the Internet years' most significant companies have been ones that made their first priority to build scale and volume. For all the above examples, their businesses are all centred on very large IT platforms and their economic models often involve selling at very low prices, or even giving services away, in order to pull in more volume.

More broadly, it's arguable that the history of IT saw a long phase -- from the 1950s to the 1990s -- in which first mainframe and then LAN computing benefited big firms, and the bigger the firm, the greater the productivity boost. IT reduced the cost of aggregating information, and therefore increased the maximum efficient size of organisations; it also permitted greater internal specialisation and the Taylorisation of many kinds of commercial and clerical processes. If you are a supermarket chain, a world-wide container shipping line, a government, a bank, or a car manufacturer, it was great. You had the capital to build the sort of IT system that made all the distinctive innovations of Japanese management practical: just-in-time delivery, lean production, global supply chains, and kanban tracking.

Spreading the benefits

Since the mid-1990s, the trend swung the other way; the Internet, and more specifically the Web, realised the potential of PCs as a decentralising technology. Small firms and individuals could now benefit; but even so, many of the real business benefits still accrued to the biggest spider on the Web, Google. Together with the other big spiders, Google had the ability to make investments in development and infrastructure big enough to create truly impressive business systems. Google aggregates vast numbers of tiny advertising opportunities up to a scale at which they are saleable. Likewise, Amazon's mammoth IT and fulfilment infrastructure lets it aggregate more titles than any other bookseller in history.

But if you are a one-truck owner operator, you still can't gather anywhere near the same amount of information Maersk or DHL can about where the next load will come from. Furthermore, you can't afford to build something that could. One of the biggest barriers for everyone is the need to invent the functions the Web doesn't have:

  • no authentication,
  • no awareness of location, presence and availability,
  • no awareness of social or legal constructs like age or jurisdiction,
  • and most of all, no awareness of money.

Market makers in information come to dominate

These absences are all opportunities for platforms to intermediate transactions and relationships. Each represents an information asymmetry, with buyers and sellers of data needs to find each other. Historically, highly profitable institutions have grown up to permit large numbers of small enterprises to aggregate such information and get rid of transaction costs: consider stock exchanges, banks, Internet exchange points, or GSM roaming hubs like Starhome. It is, in fact, the very purpose of a market. We also know that such enterprises display strongly increasing returns to scale -- business follows liquidity, rather as container ships followed traffic in the 1970s, creating a small number of giant load centre ports around the world.

One of the important points this raises is that scale is doubly important; not only does traffic go to traffic, but the elimination of transaction costs implies that the marginal cost of production must itself be very low. This is characteristic of mass production. Telcos were the first businesses to handle information goods using standardised flow processes and large amounts of capital goods to achieve very low marginal costs at very high scale, the characteristic features of mass production; but beyond their staple voice business, they seem unable to get beyond extremely cumbersome and expensive job-production. Everything is a one-off.

The telco platform fills the gaps around the Internet

Equally, the big systems at Google, Amazon, IBM, and Salesforce.com are one-offs, hugely expensive engineering masterpieces. But the characteristic feature of IT post-1995 has been that really gigantic scale in infrastructure brings down the marginal cost of transactions to a level at which very small enterprises can participate. This doesn't apply, however, to a whole realm of business processes where the things the Internet doesn't provide are vital. Here is the opportunity for the 2-sided telco platform. Telcos have key capabilities that directly address those four Internet deficits we mentioned, by growing out of their existing assets and customer bases. The purpose of a telco is to facilitate communications.

The question is how to make these available in a readily usable form; the platform is the answer. Operators aren't product innovators, and few will ever become successful media companies. They job is to re-package their key capabilities: transporting data, enabling payments, removing trust barriers, and offering B2B value-added services. These are sold via a platform business, which only works if you get three things right:

  • Create appropriate technical APIs, which match the platform user's problem. Don't offer a location dip API, do offer a service to delivery companies looking to schedule parcel drop-offs only to occupied homes.
  • Offer commercial terms of business in a transparent and efficient manner. Do not expect the platform's users to go bizdev deals with hundreds of carriers to get global coverage.
  • Offer organisational and operation systems. Sales, training, education, care, professional services and technical support.

Removing the friction from the digital economy's supply chain

Telecoms companies are in a position not just to compete with Big Tech, but to address a whole swath of the economy that falls beneath Big Tech's minimum efficient scale; trading with both sides of billions of transactions, reducing the cost and friction for both parties and taking a tiny slice of each. This may seem utopian, but emerging platform businesses like Amazon are already moving towards this. Amazon's portfolio of services now includes the ability to sell through their portal, to deliver through their infrastructure, to store information in their servers, to process it through EC2, to organise customer-support workflow through Mechanical Turk, and to bill for it through DevPay. Watch out! But even they don't have the inherent potential that lurks in the telco databases and networks...