Are you braced for the telco silver tsunami? Who’s going to look after all that legacy infrastructure that’s not going to retire any time soon?

IDC says to cope, enterprises must find creative ways to hire, train, upskill, and reskill their employees, and that somehow fostering a “learning culture” is the way forward. Well, maybe easier said than done… and what so many of these conversations ignore is what is surely an equally important pillar of modern economies, telco.

The context: we’ve been concerned for decades about skills gaps in IT and telecoms, but most of that conversation has been about just getting enough warm bodies to man the workstation or set up the routers; in the US, it’s claimed that demographic changes mean across all sectors, a ‘silver tsunami’ of retiring Baby Boomers will lead to consequences we can’t really foresee. In fact, 86% of US-based CIOs surveyed in 2023 predicted that a minimum of 25% of their network engineers will retire in the next five years.

Foresee, maybe: worry about, yes. With ongoing restructuring in many parts of our industry telcos, many of us are either being edged out whatever our ages say on our driving licences or may be induced to leave earlier than official retirement ages. Some, says PwC, will move on for other reasons, like scepticism their companies will even be around (‘I don’t think my company will still be in business within the next ten years’; 46% of telco respondents versus 31% of other industries).

How will we keep the metaphorical network light on?

But it’s probably our own silver tsunami that’s the biggest problem.  As generations of greybeards leave the org chart and all that accumulated knowledge disappears to the golf course/potting shed/easel or however you plan to spend your last decades, it isn’t just shiny new AI where there will be a skills gap — we won’t have any one around for less glamorous but still absolutely necessary parts of the stack, like maintaining and evolving intelligent network services.

Those things like emergency call distribution, operator enquiries, roaming, and non-geographic number translation that have been around for years, and are part of the furniture, really — but which are also tightly regulated and even mandatory services that we expect to be around for the next decade or so. 

To work, networks depend on a range of mandatory services that we built and want to maintain; unless we are prepared to rip them all out and somehow start again (hint: who would pay), we don’t want these less glam but still absolutely vital parts of infrastructure to just stop working.

While hardly the first words a recruitment consultant would try and tempt you with, these services need a clear pathway to the future and need to be maintained in the meantime. Take Signalling System No. 7 (SS7), the telephony signalling protocol family developed in the 1970s that was (and, in some places, still is) used to setup and teardown telephone calls – and which is responsible for enabling those intelligent network services to which we referred earlier, as well as a number of other vital functions. Even in their heyday, knowledge of these areas was in short supply, and is now scarcer still. Who's going to do this work? Young ‘uns? 

We've also met leading analysts who gained their experience in the late 90s / early 00s and who have never heard of things like SS7, so the knowledge is already disappearing rapidly. They weren’t even taught about it in their telecoms engineering studies (which may say more for the university concerned). After all, SS7 is still integral to roaming; if you don’t know how that works, what kind of telecoms expert are you?

While individual operators may know this and are already trying to work out what to do next, it doesn’t seem to be a problem our industry is as aware of compared to other parts of hi-tech.

For instance, STL Partners recently published an update to its Future Skills tracker, which assesses penetration and recruitment of nine key skills across leading telcos' and techcos' employee bases. What it showed is that, just as with regular tech companies, we’re getting very excited about things like skills in AI, Machine Learning, ‘innovation’, cloud, ‘software’. 

Great—but what about the things we’ll still need in five, ten years to keep the lights on but which are a million light years away from any conversation at a university employment Open Day? Are we risking continuity, in other words, for ambitions to create value in new areas we don’t really know how to do yet (viz operators telling McKinsey they want to see ‘digital, automation and AI at scale’ as the main way forward)?

Why would that AI postgrad work for you instead of some funky startup?

The problem here is that the same answer’s coming from our peers in healthcare, banking, retail… everyone is chasing the same shrinking pool of grads (though we could, and probably must, look at expanding the catchment for younger workers beyond the developed world anyway), and not all Gen Z want to work in anything like the kind of role that you expect, anyway.

OK; so, if we can’t get fresh blood in to take on the SS7 work (as it were), what can we do? Will we end up like CIOs who have to pay for ancient but still very much in-demand septuagenarian COBOL programmers to drive their Porsches up and patch up their equally ancient spaghetti code, lest the machines stop?

Short answer: yes, you should probably be preparing for that and quietly setting aside budget. We also wouldn’t put too much hope in AI coming in instead and doing the work for you, at least in the sense of straight replacement. What seems to be happening with AI and coding is that it can (and is) replacing a lot of routine work, especially in testing, but it is simply not good enough (yet?) to completely substitute for experienced humans.

Also, let’s be honest, if there is any AI out there in the wild that could help telcos, it’s going to be LLMs that have been trained on 5G, not boring but still vital legacy tech. But an area we think AI could and should be employed is in what a very different area, insurance, is trying to do with its equivalent problem of specialised skills in soon-to-be-retiring heads: try and capture and duplicate them in software.

A possible workaround to the telco silver tsunami?

According to a US HR research group called The Josh Bersin Company, this so-called digital twin approach is already being trialled in several of what it identifies as leadership organisations in that sector, like Zurich and MetLife: 

“Imagine you have a superb customer service agent with years of experience helping your most demanding clients. If you load the last five years of their emails, coupled with all their internal documentation, and a long history of their last two years of service calls, you can essentially ‘create him or her’ digitally with all the knowledge, style, and internal contacts this person has developed.

“[Such a] twin, which may look initially like an AI assistant, could then carry on this employee’s work when the real-life worker is on vacation. One of our clients, a large insurance company, has already built such digital twins for claims processing. If you think about the complexity and workflow of processing a claim, much of it could be learned by an agent, making the claims robot an expert on this important process. And as you change claims rules and limits, the agent will learn new guidelines in only seconds.”

We can easily see a similar approach proving useful in the “SS7” (let’s generalise and write ‘legacy’) end of telecoms, with concerted efforts to stop vital expertise exiting the building in a few years. Let’s be honest: in the face of massive change in the way we must staff and retain our companies going forward, can a better way of maintaining this knowledge other than a vague quest for more “software” or hugely competitive “AI” skills chasing solve the ageing-out problem?