Thursday, September 21, 2017

Up the Stack, Down the Stack Or Not?

One recurring and important strategic issue was raised at both the Spectrum Futures event and the following PTC Academy training course held in Bangkok, Thailand the week of Sept. 18, 2017: Can service providers move up the stack, and should they?

Russell Lundberg, Bangkok Beach Telecom CEO and founder, argued that “I’m a plumber; I can’t worry about moving up the stack.” Instead, his view is that service providers must “embrace their dumb pipes.”

Allan Rasmussen, Yozzo Co. managing director, took the other position. “You can move up the stack, but you must partner” to do so.

A third position was offered by Marc Olivier, Sigfox VP, namely that many new business opportunities in the internet of things area--especially for machine-to-machine sensor apps--are best handled by networks optimized for such applications. Olivier leaned towards the “stick to your knitting” approach, pointing out that a new ecosystem has to be built.

John Kjellemo of Yandex also provided his views about the internet of things, and believes that moving up the IoT stack is possible, and desirable.


At the PTC Academy training event, session facilitator Chris Wilson, Time Dotcom Bhd. CEO of Asia, suggested that, whatever the merits, “few companies have ever succeeded at doing so.”

My own concluding argument at Spectrum Futures was that as difficult as it is to move up the stack, retail service providers serving business and consumers really do not have much of a choice. If one assumes half of all current revenue will be lost over a decade (every decade, in fact), huge new sources of revenue must be generated, and it is hard to see how that can be done any other way.

To be sure, not every actor, in every industry segment, has equal ability to do so. So the admonition to move up the stack is not useful advice for every firm, in every part of the ecosystem. “Not moving up the stack” makes good business sense for some companies, in some industry segments.

The trick is partly knowing whether such strategies make sense, and knowing where to look for opportunities to move up the stack.

Separately, “moving down the stack” also is a big trend, exemplified by firms such as Google becoming a device manufacturer, undersea network operator, retail mobile operator and internet access provider in the fixed networks realm.

These are asymmetrical challenges. Though one frequently hears the refrain that “telcos cannot innovate,” in truth, large firms often find it hard to innovate.

But there is a structural reason why moves up the stack are so much harder than moving down the stack. A telco has to work very hard to identify what a customer wants and needs, up the stack. Much knowledge of business processes is needed.

When a company at the application layer wants to move down the stack, the problems are simpler. Any company at the top of the stack “is the customer.” That firm knows exactly how it operates, where advantage down the stack might lie, and why business advantage can be gained.

The task then is simpler: identify the capabilities needed in the lower levels of the stack, and buy them. With enough scale, it also is possible to build such capabilities, without buying them on the open market. That is why many content and application firms build and operate their own undersea networks; access networks (fixed and mobile); data centers; create their own chipsets; create their own messaging and voice capabilities and take other steps to vertically integrate their supply chains.

No comments:

"Tokens" are the New "FLOPS," "MIPS" or "Gbps"

Modern computing has some virtually-universal reference metrics. For Gemini 1.5 and other large language models, tokens are a basic measure...