{ "source": "Cooked Index — occupational AI risk register", "page": "https://cookedindex.com/jobs/market-research-analysts-and-marketing-specialists/", "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": "Market Research Analysts and Marketing Specialists", "soc_code": "13-1161", "category": "Business", "us_employment": 899580, "median_annual_wage": 78760 }, "verdict": "COOKED", "risk_resistance": 28, "contested": false, "near_boundary": false, "dimensions": { "task_resistance": 6, "embodiment": 4, "liability_shield": 2, "trust_premium": 7, "judgment_accountability": 9 }, "reasoning": { "task_resistance": "Writing Likert batteries, running cross-tabs and significance tests in SPSS or Qualtrics, scraping Nielsen/Mintel category reports into a competitive summary, and assembling the monthly campaign performance deck are all text-and-table transformations an LLM plus a BI connector now does end-to-end, which is why this sits at 6 rather than mid-band — the residual manual work (recruiting and moderating a focus group, negotiating a panel vendor's incidence rate) is a minority of the modal week, not half of it.", "embodiment": "A 4 rather than a 0 reflects the occasional in-person duty — intercept interviews at a store entrance, sitting behind the glass at a facility, shop-alongs and IHUT setup — but the deliverable is produced entirely on a laptop, and most fielding is now online panel.", "liability_shield": "No state licence, no board, no statutory signature; the Insights Association's PRC credential and the MRS certificate are resume decoration that no client contract requires, and when a segmentation study leads a brand into a failed launch the exposure is the agency's contract and the CMO's job, not the analyst's — hence 2, not the 5-10 band where certification actually gates the work.", "trust_premium": "A 7 acknowledges that a client-side research lead builds a real multi-year relationship with a brand team and gets called before decisions are framed, but the modal analyst's output is a deck circulated under an agency logo and RFP'd against two competitors on cost per complete, so the relationship travels with the account, not the individual.", "judgment_accountability": "The 9 covers genuine discretion — deciding a low-incidence B2B sample doesn't support the claim leadership wants, choosing a k-means solution over latent class, calling a lift result noise rather than signal — but the analyst recommends and the brand or product owner decides, and much of the day runs on established methodology and tracker rules rather than unprecedented calls." }, "rationale": "The modal worker in this title spends the day writing survey instruments, cleaning and cross-tabbing response data, pulling competitor and category summaries, building dashboards, and turning it all into slide decks and campaign performance write-ups — exactly the text-and-table work generative models already do at usable quality and near-zero marginal cost. There is no license, no signature, no physical component, and the client rarely buys the person rather than the deck. What survives sits in the thin upper tier: framing which question is actually worth answering, moderating live qualitative work, and standing behind a recommendation that reallocates real budget.", "outlook": "Expect headcount to compress sharply at the junior report-producing tier while a smaller senior cohort of research designers and executive-facing advisors keeps its pay and gains leverage.", "what_would_raise_it": { "levers": [ { "dimension": "task_resistance", "change": "Task-mix shift: once cross-tabs, dashboard building and deck drafting are fully automated, the residual role is research design (what question is worth asking, sampling frame validity) plus live qualitative moderation — IDIs, focus groups, ethnography — where respondent rapport and real-time probing are not yet substitutable. This raises task_resistance only for the minority who actually hold those tasks; the modal cross-tab analyst has no such tier to retreat into.", "plausibility": "already happening", "would_add": 3 }, { "dimension": "task_resistance", "change": "If synthetic-respondent panels (Synthetic Users, Evidenza, and similar) get publicly caught producing a bad category read, buyers may demand human-verified sampling and fielding audits — bot-detection, panel fraud review, incidence validation — which is judgment-heavy nuisance work AI cannot self-certify.", "plausibility": "plausible", "would_add": 2 }, { "dimension": "liability_shield", "change": "FTC substantiation enforcement is the only live route: if advertisers face penalties for unsubstantiated claims backed by synthetic or AI-generated survey data, a named human researcher may have to attest to methodology in the substantiation file. Analogous to the existing requirement that Nielsen/MRC-accredited audience measurement be human-audited (MRC accreditation already requires named auditors).", "plausibility": "plausible", "would_add": 3 }, { "dimension": "liability_shield", "change": "If state privacy regimes (CPRA, Colorado AIA) or the EU AI Act's profiling provisions require a named human to sign off on data protection impact assessments for consumer segmentation models, some analysts inherit an attestation duty.", "plausibility": "plausible", "would_add": 2 }, { "dimension": "trust_premium", "change": "Narrow route only: Insights Association / ESOMAR certification plus a client-side procurement rule requiring disclosure of whether respondents were human. Some buyers already write 'no synthetic respondents' into MSAs. This buys a premium for the fielding and qualitative side, not for the analyst who writes the deck.", "plausibility": "plausible", "would_add": 2 }, { "dimension": "judgment_accountability", "change": "If the role is redefined so the analyst owns the go/no-go on product launch or budget reallocation rather than supplying inputs to a marketing director who owns it — i.e. absorbs a slice of the brand-manager decision as that layer thins — accountability rises. Depends entirely on org design, not on the occupation itself.", "plausibility": "plausible", "would_add": 3 } ], "ceiling_note": "Even with every lever firing, this stays in the 40s. There is no license to capture, no physical component, and the deliverable is text and tables. The liability and trust routes attach to fielding and methodology attestation — a small subspecialty — not to the ~900k modal workers who cross-tab and build decks. Nothing here rescues that majority." }, "adjudication": null, "employment_history": { "points": [ { "y": 2017, "emp": 596450, "wage": 63230 }, { "y": 2018, "emp": 638200, "wage": 63120 }, { "y": 2019, "emp": 678500, "wage": 63790 }, { "y": 2020, "emp": 690160, "wage": 65810 }, { "y": 2021, "emp": 727540, "wage": 63920 }, { "y": 2022, "emp": 798620, "wage": 68230 }, { "y": 2023, "emp": 846370, "wage": 74680 }, { "y": 2024, "emp": 861140, "wage": 76950 }, { "y": 2025, "emp": 899580, "wage": 78760 } ], "from": 2017, "to": 2025, "change_pct": 50.8, "comparable_from": 2019, "spans_soc_revision": true }, "pivots": [ { "slug": "marketing-managers", "title": "Marketing Managers", "verdict": "EXPOSED", "risk_resistance": 40, "median_wage": 166790, "overlap": 80, "skills_to_close": [ "Management of Financial Resources", "Operations Analysis", "Social Perceptiveness", "Negotiation" ] }, { "slug": "fundraisers", "title": "Fundraisers", "verdict": "EXPOSED", "risk_resistance": 42, "median_wage": 72550, "overlap": 77, "skills_to_close": [ "Negotiation", "Management of Financial Resources", "Persuasion", "Coordination" ] }, { "slug": "securities-commodities-and-financial-services-sales-agents", "title": "Securities, Commodities, and Financial Services Sales Agents", "verdict": "EXPOSED", "risk_resistance": 51, "median_wage": 78660, "overlap": 77, "skills_to_close": [ "Management of Financial Resources" ] } ], "license": "https://cookedindex.com/terms" }