Job family
Assessing administrative and business support
Roles that hold an organisation's schedule, inbox and records together on behalf of somebody else.
This is the largest hiring family in the published statistics and one of the least well screened. Office and administrative support is 18.79 million US jobs, and secretaries and administrative assistants alone generate roughly 314,400 openings a year — more than customer service representatives — even while the occupation is projected to shrink 2 percent. As with several large families, the decline and the hiring volume are not in tension: automation removes the routine keying while turnover keeps the requisitions open. The one segment BLS projects to grow is medical, at 979,000 jobs and 5 percent, alongside adjacent non-clinical healthcare administration such as medical records specialists (200,700 jobs, 8 percent growth). Those vertical segments are covered by their own families — `healthcare-administration-coding` and `legal-operations-paralegal` — because the documentary and regulatory rules they work inside change the assessment. What is left here is the generalist core: administrative and executive assistants, operations coordinators, office and business support.
The hiring pain is that the category is enormous and the failure mode is quiet. The BLS projects roughly 1.7 million openings a year across office and administrative support occupations, nearly all of it replacing people who leave rather than net growth, which means employers are running the same screen over and over at high volume. A bad administrative hire does not fail visibly in week one. They fail by degrees: the meeting that was booked without the room, the email to a client that was technically accurate and tonally wrong, the expense report that was submitted with the wrong cost centre and quietly corrected by someone senior every month for a year. By the time the cost is legible it has been absorbed into three other people's workloads.
The reason these hires go wrong is almost never capability and almost always judgment about priority. An executive assistant's real skill is deciding which of four things that all claim to be urgent actually is, protecting a principal's attention without becoming a bottleneck, and writing to a senior external party in a voice that is not their own. None of that is visible in the "organised, proactive, detail-oriented" language that dominates both the job ads and the CVs answering them, and a reference check will not distinguish it either. In practice the difference is legible in about twenty minutes of simulated work: a strong coordinator reads five competing requests, correctly identifies which one is genuinely urgent versus merely loud, handles three, escalates one with a recommendation attached rather than a question, and writes the fourth up so the recipient can act without a follow-up. A weak one processes them in arrival order and asks their manager about all five.
This family carried `savvanta_fit: medium` through Phase 1 and now carries `high`. The rating was revisited deliberately, because it governs how much work the largest hiring family in the corpus gets, and the two things the medium rested on have both since fallen away. The first was the physically present half of the work — reception, front-desk, facilities. That is now excluded from this family by name in the taxonomy's own scope list, and no remaining title in `typical_titles` carries it; BLS itself reports administrative services managers (279,000 US jobs in 2025) separately from facilities managers (165,900), so the line is drawn in the source data and not only in our copy. The second was an inability to predict whether someone is still good at month six. That is true, and it is equally true of every family in this corpus — of a support agent, a collections advisor, an accounts payable clerk — so it cannot be the reason this one ranks below its siblings. A caveat that applies everywhere ranks nothing.
The rating that survives scrutiny is high, and the discriminating test is what the corpus's other medium families have in common. Machine learning, operations management and product each sit at medium because the unit of real performance is a multi-month arc — a research programme, a delivery quarter, a two-year roadmap — and a session can only sample a proxy for it. Administrative work is the inverse case. Its unit of performance is a day, and its output is literally the artefact type Savvanta produces: the inbox triaged this morning, the calendar rebuilt around a cancelled flight, the careful written no that went to a client at five o'clock, the records task where two systems disagree. A twenty-minute simulation here is not a proxy for the job; it is a sample of it, in the same sense that a simulated call is a sample of a support agent's job. That is the standard the high-fit families are held to, and this family meets it more directly than most. The honest residual limit is not a rating issue but a deployment one, and it is sharpest for executive assistants: that hire has a legitimate component of personal fit with a specific principal, which no assessment should claim to measure and which a strong score does not settle. Rank ability with the simulation; decide fit with a human.
The gain over the status quo is correspondingly large, because the status quo is a CV, a conversational interview and sometimes a typing or Microsoft Office multiple-choice test. The conversation rewards people who talk well about organisation, which is a different skill from being organised; the software test measures whether someone knows where the pivot-table button is, not whether they would notice that the numbers in the spreadsheet do not reconcile.
Why this work can be assessed
Most of the output is written — emails, briefs, records, calendar decisions — and the hard part is prioritisation under conflicting demands, which a judgment scenario with a deliberately over-full inbox reproduces well.
Roles in this family
Sources
Every figure on this page is traceable. Where a claim could not be sourced it is stated qualitatively instead.
- US Bureau of Labor Statistics, Employment Projections, Employment by Major Occupational Group 2025-2035, https://www.bls.gov/emp/tables/emp-by-major-occupational-group.htm
- US Bureau of Labor Statistics, Occupational Outlook Handbook, Office and Administrative Support Occupations, 2025-35 projections, accessed 2026, https://www.bls.gov/ooh/office-and-administrative-support/home.htm
- US Bureau of Labor Statistics, Occupational Outlook Handbook, Secretaries and Administrative Assistants, 2025, https://www.bls.gov/ooh/office-and-administrative-support/secretaries-and-administrative-assistants.htm
- US Bureau of Labor Statistics, Occupational Outlook Handbook, Medical Records Specialists, 2025, https://www.bls.gov/ooh/healthcare/medical-records-and-health-information-technicians.htm
- US Bureau of Labor Statistics, Occupational Outlook Handbook, Administrative Services and Facilities Managers, 2025-35 projections, https://www.bls.gov/ooh/management/administrative-services-managers.htm
- US Bureau of Labor Statistics, Occupational Outlook Handbook, General Office Clerks, 2025, https://www.bls.gov/ooh/office-and-administrative-support/general-office-clerks.htm
Hiring for one of these? We build the assessment for the specific role, run it under your brand, and return a ranked list with the evidence behind every score.
Book a walkthrough