Last week I made the (perhaps) bold statement that I do not consider autonomous driving a consumer 5G use case.

Like home security and other similar services, consumers will not be paying a mobile operator directly for the connectivity that autonomous driving requires. They will have paid more for their car and may be paying a subscription to their car manufacturer. Or none of us may own cars and Uber or Lyft will operate autonomous ‘taxis’. Or, as appears to be the thinking in some countries, autonomous driving is a road infrastructure issue and we will be paying for it in tolls or through our taxes. But in whatever circumstance, the infrastructure for autonomous driving will be justified and paid for by B2B revenues, not consumer revenues – except indirectly.

As I said last week, some may see this argument as ‘splitting hairs’. There is, after all, a consumer paying for something.

But that is always true. The iron ore leaving a mine in Australia eventually becomes some piece of consumer gear – like a car – and so is, somewhere at the end of the value-chain, ‘consumer’. In fact, I believe everything, every value chain, eventually has a consumer paying for something (maybe directly like a car or maybe in the form of taxes like government services).

A more important argument is Why do we care about the distinction? Why does this matter? It mattered last week because I was making an argument about Nokia’s Enterprise strategy. But does this matter really?

I think it does.

The channel is different. Selling to millions of consumers is a completely different proposition from selling to less than a dozen car manufacturers. So is servicing. The party with the contact center to service millions of vehicles will be the auto manufacturer or the highways department, not the telecom operator.

I think the architecture will be different as well. Or perhaps, stated differently, the architecture will not favor mobile operators. As I understand how autonomous driving will work, there will be a lot of communication between cars and between cars and infrastructure. Yes, there will be wide-area communications for maps, traffic updates, weather updates and other types of information necessary to make strategic decisions like getting off the highway now to avoid a flood or a traffic jam. But high-value tactical decisions like crash avoidance or lane changes needs low latency communications which will be short range, probably in some kind of mesh architecture. I cannot imagine lane changes or crash avoidance decisions being hauled back to some central point for any purpose. The speed of light is an unavoidable constraint.

I struggle to come up with a high-value, essential role for a traditional mobile operator with this kind of network.

This architecture and the demands of autonomous driving may mean that auto manufacturers will not want each of their drivers making individual operator choices. The design parameters may require a high degree of confidence about when all vehicles in the same area receive the same information. That may require all to be on the same network, at least for the wide-area communications part of the system.

A major point is that the margins are different with this service model. In a B2B2C value chain, normally, the value is at the B2C end and margins are higher where value is perceived to be higher.

The B2C end is, at the very least, where the perception of value occurs – I am buying a car or buying transportation and that is what I find of value. I may need wide area communications to accomplish autonomous driving but, in the worst case, I am completely unaware of that. Thus, I may not ask too many questions about prices for something where my perceived value is high. However, if you ask me to pay separately for wide-area communications, even for autonomous driving, I will probably do price comparisons.

That said, individual drivers would have less bargaining power than a huge, usually global, auto manufacturer or a department of highways.

To pick but one part of a more complex discussion, if an individual threatens to leave an operator, the MNO has a relatively easy decision to make: is what I need to protect this renewal more or less than the profit I will earn from this customer over the contract period. The risk of making a bad decision is one ARPU. Better still, the majority of clients will renew automatically without even thinking about negotiating a better deal.

However, if General Motors wants a better deal at renewal time, the risk is losing millions of connections in a single blow. If each connection was on an individual contract, perhaps 90% or more would never consider changing providers. But it is virtually guaranteed that General Motors will shop around and ask the question about whether tariffs could be lowered.

The consequence has to be lower margins in the B2B segment.

Bottom-Line: Most operators say that the 5G business case is challenging, mostly because the essential use cases – those which can only be realized with 5G – are, today, theoretical.

Only the consumer-oriented eMBB (enhanced Mobile Broadband) and FWA (Fixed Wireless Access) use cases are based on analogies to services that we have today. The enterprise business cases, including autonomous driving, are largely uncertain: will there be enough uptake and revenue to justify building out a 5G network?

All business cases have inherent risks but those which are analogous to what we have today have several advantages: marketing research techniques like surveys are more effective, successful marketing strategies are already well established and it is easier to build forecasting models based on migration and growth. Furthermore, Boards of Directors will feel more comfortable with decisions where the knowns seem to be more than the unknowns.

As one network planner at a multinational operator told a small group of analysts once, “If we do not have a strong consumer business case, it will be hard to get 5G past the Board.”

Which is the main reason why I think the distinction between consumer and enterprise is important. Boards understand (or perhaps think they understand) consumer markets. B2B or B2B2C business models are less familiar and so harder to sell when it comes to justifying the Capex.

In the case of the US operators, eMBB and FWA were ‘enough’ to get the Board to agree to billions of dollars in Capex. Scarce spectrum probably played an important role in justifying the additional investment.

It is not clear if these will be enough for other operators in other countries, especially in Emerging Markets.

Title Reference: From the Beach Boys’ 1963 song “Little Deuce Coup” from the album of the same name. It was the B-side to the (perhaps better known) single Surfer Girl. I picked it as a car reference but also because of the concept that we don’t really know much about 5G business cases, especially autonomous driving.

 

No Comments »