Thursday, October 5, 2017

Pick Your Poison: Smaller Markets or New Competitors

Amazon.com has launched its own delivery service, after already taking a stake in an  air freight company (Air Transport Services Group) and creating a branded “Prime Air” service. The launch might be characterized as a test, but that might also have been said of the Prime Air service, which has expanded. Amazon also has built its own air freight hubs.

That investment aims to vertically integrate at least some of Amazon’s own long haul operations, by creating its own internal logistics network for air freight.

The latest move goes further, and aims to create an Amazon-owned delivery service direct to its customers. In large part, the move aims to speed up delivery options, but also free up space in Amazon warehouses, as goods would move directly from manufacturer locations to distribution, bypassing delivery companies such as United Parcel Service  and FedEx.



Some of you already will have understood the similarity to trends in the global telecom business, where enterprises that formerly were “customers” actually vertically integrate their own long haul or access functions as well as data center operations (fixed internet access or mobile access).

That removes demand from the public telecom markets. It is not so much that the enterprises become competitors as that they cease to be potential customers. Some might see an enterprise that builds and operates its own communications networks as a “competitor.”

In most ways, it is worse than that. Competitors represent a threat to revenues and profits, to be sure. But there always is the possibility to reclaim market share from such firms.

When enterprises create their own facilities-based networks, they remove demand from the markets all the competitors serve. In part, that limits the size of markets, as some amount of present revenue and future growth is removed.

It is not an unusual story. In many ways, cloud computing internalized by enterprises eliminates demand for data center services. Other trends, such as open source, also limit sales and growth  in other areas such as servers and associated gear.

None of this is unusual, as a business strategy. Telcos often speak about “moving up the stack” into the applications layer. In industrial value chains, “backward integration” is the equivalent of a service provider moving further “down the stack.”

Forward integration, in an industrial value chain, normally describes occupying new niches in the value chain (towards retail and away from raw materials) that correspond to a telco “up the stack” strategy.

If, as many expect, enterprises in the 5G era are more able to create their own mobile and wireless networks using a mix of unlicensed and shared spectrum, as well as licensed spectrum, then two things will happen.

In some cases, end user demand will be removed from the market. In other cases, former enterprise customers could become actual direct competitors to service providers. Pick your poison.

No comments:

Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...