Thread by @erlendpre

The telecom industry seems perpetually surprised that other industries keep taking a piece of the revenue they tradtionally had. Smartphones, messaging apps, sea cables, roaming etc etc... Now it will be networks themselves.

Not the big macro networks, but the smaller networks. Unlicensed spectrum, 5G, networks as a service will drive 10-100x the number of mobile networks in just a few years. This in turn will redefine how macro networks will eventually look.

The thing is, it isn't a out the technology (mainly). Operators have access to much of the same tech as everyone else. It is however about technology control, development, operations - and perhaps most importantly, technology closeness.

By having outsourced so much of their technology, operators have lost the deep technology understanding necessary to help define the next wave of change. Others step in their place.

Observe now the traditional tier 1s taking network as a service to new levels, device manufacturers bypassing operators, and a raft of new private network companies coming into play. Common to most of them is that they develop and operate their core technology.

Keep a very close eye on operators taking conscious choices on this. You will find a polarization I think. Some operators will be the "Gillette" omin their country. Brand, marketing and distribution companies. Some will be taking a position as programmable networks for others.

A very few will take on the ambition to themselves become tech companies.

What will drive the most value? Hard to say, the last category has the most potential, but is the hardest to do.

@proximus is the company I have seen with the clearest ambition of being a programmable network. Jio, Rakuten and a few others take more tech in-house.

The hyperscalers are a bit of a joker here. By being the underlying infra providers they will (at least) partly deliver the flexibility and programmability the operators require, essentially replacing the traditional tier 1 vendors.

By building infra across a number of industries, the hyperscalers are now at capex levels that are insane. Operators will struggle to match the quality, distribution and performance of this infra.

Then of course there are hurdles to leveraging hyperscalers. Regulatory constraints, certain performance requirements, national autonomy, pricing and more - but more likely than not, these are challenges that with time will be solved.

Also interesting to witness hyperscaler positioning. Azure now offering the infra, and the applications (metaswitch/affirmed). Google unclear to me still. AWS takkng a platform approach.

So, the future involves a 100x the number of networks, hyperscalers with 10-100x the infra of anyone else. And with 100x the number of companies and use cases leveraging mobile networks. What does it mean?

I am reasonably sure that imperatives are i) ability to leverage hyperscalers, ii) drive marginal cost of deployment and integration close to zero and iii) drive programmability of networks to drive marginal cost of operations down

Operators are important, and fulfill critical functions in society. They are also one of the few truly global and highly distributed industries on tech. I would love to see them take a bigger part of the future of tech as it would disaggregate power. But it won't happen...