Job family

Assessing it support and service desk

Roles that fix other people's technology problems, over and over, at volume and against a clock.

This is the highest-volume family in technical hiring and the one screened with the least rigour. BLS counts 903,100 US computer support specialists in 2025 — one of the largest occupations in its computer and information technology group — with roughly 48,700 openings projected each year despite a 3 percent decline in total employment — the churn, not growth, is what drives the hiring. Median pay of $62,890 overall, rising to $76,220 for network support specialists, puts this squarely in the volume-recruitment bracket where employers process hundreds of applications per opening and have no economical way to look closely at any of them.

The work is a diagnostic conversation with someone who cannot describe their own problem accurately, conducted against a service level agreement. What separates the top quartile is the shape of their questioning: they establish what changed before they establish what is broken, they distinguish the symptom the user reported from the fault actually present, and they know when the honest answer is to escalate rather than keep the user on the line proving competence. The expensive failures are behavioural rather than technical — the technician who closes the ticket without confirming the fix, who resolves the issue but leaves a note nobody else can act on, or who is technically correct and leaves the user feeling stupid. In a managed service context each of those is a renewal risk.

Status quo screening is a CV keyword filter for a vendor certification, followed by a phone screen that is mostly a personality impression, occasionally with a multiple-choice technical quiz attached. None of it observes the candidate doing the thing. The quiz tests recall of facts that are one search away on the job; the phone screen assesses how the candidate talks about support work rather than how they perform it. And because the applicant pool is large and largely undifferentiated on paper, hiring managers end up selecting on proxies — articulateness, familiarity, confidence — that correlate with background at least as much as with capability.

Savvanta's fit is high and unusually direct. A live call with an AI user who gives a vague and partly wrong account of the symptom, escalating in frustration if the candidate leads with jargon, tests the actual core of the job. The ticket note written afterwards tests whether the next person in the queue could pick it up. A short judgment scenario — the request is out of policy, the user is senior and insistent — tests the boundary-holding that support organisations quietly depend on. Because the family hires at volume, a defensible ranked shortlist with transcript evidence is worth more here per hire than almost anywhere else in the technical estate.

Why this work can be assessed

The job is a diagnostic conversation followed by a written record — a live call with an AI user who describes the symptom badly, plus the ticket note that follows, is the job almost exactly.

Sources

Every figure on this page is traceable. Where a claim could not be sourced it is stated qualitatively instead.

1. US Bureau of Labor Statistics, Occupational Outlook Handbook, Computer Support Specialists, 2025, https://www.bls.gov/ooh/computer-and-information-technology/computer-support-specialists.htm
2. US Bureau of Labor Statistics, Occupational Outlook Handbook, Computer and Information Technology Occupations, 2025, https://www.bls.gov/ooh/computer-and-information-technology/home.htm