{
  "source": "Cooked Index — occupational AI risk register",
  "page": "https://cookedindex.com/jobs/sales-and-related-workers-all-other/",
  "methodology": "https://cookedindex.com/methodology",
  "notice": "Verdicts are re-examined as evidence accumulates. Re-fetch before relying on this; the page above always carries the current score.",
  "scored_at": "2026-08-11",
  "model": "claude-opus-5",
  "occupation": {
    "title": "Sales and Related Workers, All Other",
    "soc_code": "41-9099",
    "category": "Sales",
    "us_employment": 93180,
    "median_annual_wage": 48280
  },
  "verdict": "EXPOSED",
  "risk_resistance": 35,
  "contested": false,
  "near_boundary": false,
  "dimensions": {
    "task_resistance": 8,
    "embodiment": 7,
    "liability_shield": 2,
    "trust_premium": 10,
    "judgment_accountability": 8
  },
  "reasoning": {
    "task_resistance": "At 8, the split is real but lopsided: list building, cadence emails, quote-and-proposal assembly, and CRM hygiene — likely most of a rep's logged hours — are already being handed to tools, while the discovery call where a buyer talks himself out of a purchase and has to be talked back in still needs a human to hear the pause and change tack.",
    "embodiment": "A 7 reflects that most of the week is phone, email, and video, but this bucket absorbs event and trade-show sellers, in-home demo reps, and route-and-territory canvassers who carry samples, set up booths, and drive to the buyer's site — physical presence that shows up in some jobs here, not in the baseline one.",
    "liability_shield": "A 2 rather than 0 acknowledges that some employers require product-specific certification badges and background checks, but nothing here is a state license: no exam gates the work, no registry can be struck, and the employer or manufacturer absorbs any misrepresentation claim.",
    "trust_premium": "10 puts this squarely mid-band because named accounts, repeat buyers, and referral flow do accrue to the individual rep, but the relationship attaches to the price, the product, and the territory — buyers switch reps without switching vendors, and most of the book transfers when the rep leaves.",
    "judgment_accountability": "An 8 covers the discretion that actually exists — how far to discount inside an approved band, which objection to concede, whether an account is worth the quarter's attention — while pricing floors, contract terms, and credit approval sit with sales management and finance, so the rep rarely owns the call that can't be reversed."
  },
  "rationale": "This is a residual catch-all bucket, so the modal worker is a generalist seller — prospecting lists, sending outreach, quoting prices, demoing products, logging activity in CRM — and the prospecting, follow-up sequencing, quote generation, and pipeline notes are already the strongest use case for current AI. What survives is the part where a buyer wants a person in the room: reading hesitation, negotiating terms, and being the accountable face when the deal or the delivery goes sideways. No license protects the role, and volume-based seat counts are the first thing trimmed when AI raises per-rep throughput.",
  "outlook": "Headcount in this bucket shrinks as AI absorbs prospecting and quoting volume, leaving fewer sellers who each carry larger, relationship-heavy territories.",
  "what_would_raise_it": {
    "levers": [
      {
        "dimension": "task_resistance",
        "change": "Task-mix shift: once AI absorbs prospecting, sequencing, quoting and CRM hygiene, the surviving seat is the complex-deal tier — multi-stakeholder negotiation, custom terms, on-site demos, escalation recovery. This is a genuine two-tier occupation, but the shift raises per-seat resistance while cutting seat count, so the register score rises for a smaller population.",
        "plausibility": "already happening",
        "would_add": 4
      },
      {
        "dimension": "trust_premium",
        "change": "Buyer-side procurement rules that require a named human account representative and prohibit AI-only quoting on contracts above a threshold — visible already in some enterprise and government vendor terms and in state bills restricting automated pricing/AI-set prices (e.g. California and New York algorithmic-pricing disclosure bills). If a named human of record becomes a standard contract clause, the premium hardens.",
        "plausibility": "plausible",
        "would_add": 4
      },
      {
        "dimension": "liability_shield",
        "change": "Sector-specific licensing pulling parts of this residual bucket into regulated selling: e.g. state insurance producer or securities registration extended to adjacent product sellers, or FTC/state enforcement on AI-generated sales claims requiring a named human to attest to representations made in a quote or demo. Narrow, because the bucket is unlicensed by definition.",
        "plausibility": "unlikely",
        "would_add": 3
      },
      {
        "dimension": "judgment_accountability",
        "change": "Discount and concession authority formally delegated to the rep with a signed approval of record, as AI recommends but cannot authorize — the pattern in existing CPQ approval-matrix policies. Codified as a human-authorization gate in deal desk rules, the role owns the consequential call.",
        "plausibility": "plausible",
        "would_add": 3
      }
    ],
    "ceiling_note": "Realistic ceiling is mid-50s. The bucket is a residual SOC code with no professional body, no license, and no unified employer, so nothing can shield it uniformly; whatever survives is headcount-thinned enterprise closers, and the register score rises for the remnant rather than for the 93,180."
  },
  "adjudication": null,
  "employment_history": {
    "points": [
      {
        "y": 2017,
        "emp": 93860,
        "wage": 33890
      },
      {
        "y": 2018,
        "emp": 95690,
        "wage": 33220
      },
      {
        "y": 2019,
        "emp": 118910,
        "wage": 31820
      },
      {
        "y": 2020,
        "emp": 153180,
        "wage": 30550
      },
      {
        "y": 2021,
        "emp": 235740,
        "wage": 29570
      },
      {
        "y": 2022,
        "emp": 119490,
        "wage": 44120
      },
      {
        "y": 2023,
        "emp": 114280,
        "wage": 44610
      },
      {
        "y": 2024,
        "emp": 99070,
        "wage": 46370
      },
      {
        "y": 2025,
        "emp": 93180,
        "wage": 48280
      }
    ],
    "from": 2017,
    "to": 2025,
    "change_pct": -0.7,
    "comparable_from": 2019,
    "spans_soc_revision": true
  },
  "pivots": [],
  "license": "https://cookedindex.com/terms"
}