Amazon and eBay, following their establishment in 1994 and 1995, harnessed the connectivity of the Internet to create new, more efficient markets. In the beginning, that meant new ways of buying and selling books and collectibles; but now e-commerce is everywhere, offering customers new goods and used goods－and becoming a global force in logistics and retail. Likewise, while today's sharing-economy companies may be just out of their infancy, their services will one day be ubiquitous.

By now, most people have heard of Airbnb, the online apartment-rental service. The company has just 600 employees but a million properties listed for rent, making it larger than the world's biggest hotel chains. Of course, what Airbnb offers is different from what hotels provide; but if Airbnb offered options for, say, maid service or food, they could become closer competitors than one might initially imagine.

The Internet enables creative new business models that increase not only a market's efficiency but also the utilization of our various assets. Hundreds of experiments are being conducted. Clearly, not all of them will experience the astonishing growth of Airbnb and Uber. Some, like Rent the Runway for designer clothes and accessories, may find profitable niches; others will simply fail.

The digital platforms that act as the basis of all this e-commerce need to meet two related challenges. The first is to produce a network effect, so that buyers and sellers find one another often enough and rapidly enough to make a business sustainable. Second, the platform must create trust－in the product or the service－on both sides of the transaction.

Trust is crucial to the network effect; hence the need for two-way evaluation systems that encourage buyers and sellers to be repeat users of the relevant platform. Small players can then act in large markets, because－over time－they become known quantities. The power of these platforms derives from overcoming informational asymmetries, by dramatically increasing the signal density of the market.