Retire the hardware standard, and the technicians built around it disappear with it. That prediction gets made every time a physical standard is switched off, and it sounds obviously right. One such retirement has already run far enough to check against what the prediction promised. The check does not return what the forecast assumed.
A product is easy to declare dead. A task list is a different thing, and it is the task list, not the product, that decides what actually happens to the people who did the work. Which one is retired decides who stays employable on the other side.
The One Transition That Has Already Run Its Course
US copper and PSTN retirement is the hardware-to-IP transition with an actual paper trail: regulated, timetabled, and running now. The FCC sets the procedural rules governing that retirement: affected customers must be given ninety days’ notice before service moves off copper. The retirement is accelerating through 2025 and 2026, as DataCenterDynamics has tracked. Neither source addresses the people doing the physical work; both stay on the network.
The workforce side of that retirement is not counted anywhere in the public record. The transition is documented as a regulatory and engineering event, with no published figure for how many technicians it touched or where they went afterward.
Since the headcount cannot be looked up, the check has to run where it can actually be observed: at the level of the task. What the record does carry runs the other way: a shortage of these technicians, its causes unsettled.
Three Columns, and Only One of Them Empties
Because job titles blur where the work doesn’t, run any retirement task by task — the copper pair, the satellite dish. Three columns emerge; only one empties.
What dies. The skills that exist only because of the retired standard: aligning a fixed dish to a fixed arc in the sky, the geometry that made that a specialist act, and swapping outdoor hardware built for one product. This is the smallest column, and it should be — the part of the job most tied to the medium itself, and the least of what a working day was ever made of.
What transfers. This is the largest column, and the reason the forecast fails. Working at height and on roofs; routing cable through an occupied building; isolating a fault along a physical path; diagnosing on site with a customer standing there; the scheduling discipline that makes a first-time fix possible; and the credential that gets a stranger let into a private home. None of it is specific to the medium. All of it is slow to train.
What is newly created. Work that did not exist under the old standard: covering a home’s network room to room, configuring the customer-premises equipment that replaced the old box, verifying throughput along the whole path, and a genuinely new problem — a fault no longer on the roof, no longer confined to one company’s hardware, and belonging to nobody until someone is sent to find where it lives.
Why a Shortage and a Transition Can Both Be True
DataRemote documents a technician shortage around copper retirement: copper-plant specialists are leaving the trade faster than new ones arrive. The shortage it records is attrition, not displacement. Where the remaining work goes is not its subject.
Attrition and recomposition are not competing accounts; the ledger shows why both hold. The people that shortage counts hold both columns at once: the copper-specific skill that is going, and the access-and-diagnosis skill that is not. DataRemote observes real thinning in the medium-specific skill — exactly what the dying column predicts. It does not measure the larger column riding alongside it, which the same departing workforce also carries and which no standard change removes. That reading turns the shortage into a demand signal for the transferring column, worth planning around now.
What the Market Does With a Column It Cannot Retrain Fast Enough
Where a skill is scarce, slow to train, and needed intermittently at scattered sites, the market’s answer is intermediation: routing the work, per visit, to whoever holds it. Staffing that skill permanently in every operator would waste the scarcity, so the market routes around it. Field Nation, a labor marketplace positioning itself as sourcing local technicians for site surveys and equipment installs during copper retirement, fits that pattern. What does Field Nation’s existence actually prove? How the work is routed — nothing about headcount.
Intermediation changes the unit of the industry. Because capacity is bought per visit, the visit itself becomes the thing that is scheduled, priced, won and lost, and an operator’s cost structure stops being a payroll question. Rented by the visit, that capacity also carries no payroll line for training it — the apprenticeship has to be funded elsewhere.
The Only Number the Transition Actually Publishes
The site visit is the unit that survives every standard change, and it is almost never priced in public. Operators treat the cost of sending someone to an address as an internal number; retirement programs get costed as network engineering, and the labor line inside them is not published. The transition therefore generates visit-shaped work whose price stays undocumented, with one narrow exception: where the visit is sold directly to a consumer, and the range has to be advertised to be sellable at all.
Indicative ranges published in South Africa by a referral network for independent satellite-TV technicians in Gauteng put three job types between R350 and R1,500 — estimates only, each installer setting its own rate and confirming the price in writing after seeing the site. The network installs nothing itself and is not MultiChoice or DStv.
The cheapest published job is the one most specific to the retired standard — realigning a dish — and the most expensive is a dish mounted and aligned from scratch with cabling, configuration and verification layered on top. The dying column sits at the bottom; the other two lift it to the top. When the hardware task disappears, what remains is the higher-value end of the day, and it still has to be paid for by somebody. What is published is the retail price of that day. The labor line inside it stays unpublished, and that price is the ledger’s public figure.
The Column Worth Training For
Three things follow from the ledger. Run it before the retirement date. Ask of each task: did the medium make it hard? It is a workforce-planning tool: read as a talking point, it teaches no one where to train next. Aim training pipelines at the transferring column: access, diagnosis, scheduling — the skills that outlast the standard being switched off. A pipeline built that way trains dispatch and diagnosis first. Treat the visit itself as the planning unit, scheduled and costed as such, because intermediation already made it that.
The same ledger applies wherever a physical standard migrates onto IP: copper and ADSL, prepaid metering, satellite television, and every consumer product that used to need its own dedicated box. Only the first column ever really empties. What survives, and what has to be paid for, is the workforce that could always go anywhere a fault is.
