Role Before Title: How to Identify What a Stakeholder Actually Does in the Decision

Who this framework is for

This framework is for account researchers, sales teams, manufacturers, exporters, and revenue-operations leaders who need to map named people without turning title data into unsupported authority claims. It is particularly valuable across countries, industries, subsidiaries, and smaller companies where titles vary or one person covers several functions. It maps a decision; it does not create a persona list or select the account ICP.

Contents

1 Separate the three facts 2 Start with decision work 3 Use the role-evidence ladder 4 Read titles as clues with limits 5 Classify actions and artifacts 6 Verify authority, influence, and access separately 7 Record a falsifiable role hypothesis 8 Illustrative example 9 Common mistakes 10 What this does not mean 11 Common questions 12 Sources and methodology

1. Separate the three facts

Role-before-title analysis keeps three statements distinct: 1 Identity fact: who the person is and the public title or function shown by a dated source. 2 Decision observation: what the person said, requested, produced, changed, introduced, approved, or accepted in the scoped purchase. 3 Role hypothesis: which buying function that evidence supports. For example: Identity fact: Ana Ruiz is listed as Plant Engineering Manager on the company's site as of August 20. Decision observation: Ana supplied the motor specification and requested a vibration test for the Plant 3 retrofit. Role hypothesis: Ana is a Known Technical Evaluator for that retrofit; formal acceptance authority remains Unknown. The first statement may be public. The second is decision evidence. The third is an interpretation. Combining them into "Ana is the decision-maker" discards scope and invents authority.

2. Start with decision work

Use functional verbs rather than departmental labels:

• defines the problem;

• builds the internal case;

• sets or interprets requirements;

• tests technical compatibility;

• evaluates operating consequences;

• requests risk, quality, legal, or commercial evidence;

• controls the required process;

• introduces another participant;

• recommends a choice;

• commits funds;

• approves, signs, or accepts within a stated limit;

• uses or supports the resulting solution. The U.S. Bureau of Labor Statistics explains that the Standard Occupational Classification assigns workers based on duties rather than titles, and the same title can map to different occupations depending on work performed. BLS, How Do I Find My SOC Code? SOC is a statistical system, not a buying-role framework. Its useful lesson is limited but important: title text can be insufficient even for classifying a person's regular occupation, so it is even less sufficient for proving a role in one purchase. Ask what decision work is present before looking for the nearest senior title.

3. Use the role-evidence ladder

The SellersGPT role-evidence ladder has five levels: Level Evidence Permitted statement 0 No identity or role evidence Unknown person and Unknown role 1 Public title or department only Possible functional clue; role Unknown 2 Indirect statement or scope-relevant public duty Inferred role with source and limitation 3 Direct observed decision action Known role contribution for the stated scope 4 Direct authority, approval, or acceptance evidence Known authority or stop condition within documented limits Levels are not scores of personal importance. A Level 3 user observation can be essential even when the user lacks authority. Level 4 applies only to the authority claim that is evidenced; it does not automatically prove sponsorship, influence, or support. Unknown evidence cannot be averaged into a medium-confidence answer. If a person sets specifications but signature authority is unverified, record Known Technical Evaluator and Unknown authority.

4. Read titles as clues with limits

Occupational sources can suggest normal duties, vocabulary, or areas of expertise. O*NET separates occupation titles, reported job titles, work activities, tasks, and work context in its U.S. Department of Labor-supported content model. O*NET Content Model Statistics Canada's National Occupational Classification likewise describes a job as tasks performed by a person and warns that its extensive job-title list is not exhaustive; classification relies on duties and responsibilities. Statistics Canada, NOC 2021 Introduction Use those sources to form questions, never to assert participation. A title can vary because of:

• company size and title inflation;

• local language or industry convention;

• centralized versus site responsibility;

• formal versus informal work allocation;

• temporary project assignment;

• matrix reporting;

• outsourced functions;

• a role combined with ownership in a small business. Titles also become stale. Record the source date and verify the account, site, and current function. A reliable title still says nothing by itself about buying intent.

5. Classify actions and artifacts

Use observed action and the artifact it affects: Observed action or artifact Plausible role hypothesis What remains unproven Builds the problem case and gains internal introductions Business Champion Funding, formal authority, continued support Owns the economic rationale or allocates funds Economic Sponsor Technical acceptance, daily influence Writes specification or requests a test Technical Evaluator Contract authority, economic sponsorship Defines downtime, training, service, or maintenance limits Operational Evaluator Final sign-off, budget ownership Requests due diligence, terms, vendor setup, or required clearance Risk & Process Validator Ownership of business need, product preference Describes actual work and acceptance experience User / Influencer Approval or procurement authority One action can support several hypotheses. Asking for a test may show technical evaluation and process validation. Keep primary and secondary roles, each with evidence. In U.S. federal source selection, FAR 15.303 explicitly separates the source selection authority, evaluation team, and contracting officer. FAR 15.303, Responsibilities That regulated model is not a private-sector title map. It demonstrates why evaluating, recommending, controlling communication, and awarding should not be collapsed into one "buyer" label.

6. Verify authority, influence, and access separately

Create separate claims: Claim Verification question Strong evidence Influence Has this person changed criteria, recommendation, or another participant's view? Dated action, accepted requirement, cited recommendation Authority Can this person approve, commit, sign, or accept within this exact scope? Delegation, approval record, tender role, direct confirmation Access Can this person provide information, introductions, or the required process route? Completed introduction, document access, scheduled review Interest Is the person attending to this decision now? Decision-specific question or action Sentiment Does evidence show support, conditions, neutrality, opposition, or conflict? Direct scoped statement or action Do not infer one field from another. A highly influential technical expert can lack signature authority. A formal signatory can be distant from evaluation. A helpful contact can provide access without being a champion. SEC EDGAR provides public access to company filings that may identify officers, segments, facilities, governance, and material risks for U.S. reporting companies. SEC, Search Filings A filing can establish a public corporate fact within its reporting scope. It does not prove that a named officer participates in a specific purchase.

7. Record a falsifiable role hypothesis

Use this template: For [decision scope] as of [date], [person/team] is [Known/Inferred/Unknown] as [functional role] because [specific evidence]. This does not establish [unproven fields]. The hypothesis would be weakened or disproved by [contrary evidence]. Next validation: [bounded action]. Add source, owner, and review date. Examples of disconfirming evidence include a current approval matrix naming another authority, the person stating that the work belongs to another site, or a tender assigning evaluation elsewhere. A hypothesis that cannot be disproved is a stereotype. "Operations leaders care about efficiency" is too broad. "The maintenance manager owns the shutdown plan for this retrofit because she issued the dated outage window" is testable.

8. Illustrative example

Illustrative example - explicitly fictional. Cobalt Seal Systems, a fictional component exporter, reviews six people at a fictional OEM for one new-model qualification. It assigns evidence-ladder levels to role claims, not to people. Person/title clue Observed decision work Role claim Level Director of Engineering Listed on company page only Technical Evaluator 1 Commodity Manager Requested vendor forms and terms Risk & Process Validator 3 Product Engineer Issued drawing and test requirement Technical Evaluator 3 Program Manager Introduced engineering and quality; summarized launch risk Business Champion 3 Vice President, Operations Mentioned in a filing; no decision action Economic Sponsor 1 Quality Engineer Named by program manager as test reviewer Risk & Process Validator 2 There are six role claims. Three have direct observed decision actions at Level 3, one has indirect evidence at Level 2, and two have title clues only at Level 1. Direct-action support is therefore 3 / 6 = 50%. Claims with at least indirect support are 4 / 6 = 66.7%. Neither calculation predicts approval. Cobalt records the Product Engineer as a Known Technical Evaluator and the Program Manager as a Known Business Champion. It keeps the Director of Engineering and Vice President roles Unknown. Its next action is to ask who owns economic approval and final quality acceptance. It does not contact the two senior titles merely to improve a percentage.

9. Common mistakes

• Treating title verification as role verification.

• Calling the most senior person the Economic Sponsor.

• Assuming procurement owns technical or business acceptance.

• Converting responsiveness into influence or authority.

• Using an occupational profile as proof of a named person's decision work.

• Ignoring title staleness, site scope, or matrix responsibility.

• Combining influence, authority, access, interest, and sentiment.

• Writing role hypotheses that cannot be disproved.

• Contacting a person because the title looks important rather than because a role purpose exists.

10. What this does not mean

Role-before-title does not mean titles are useless. Titles can locate expertise, suggest vocabulary, and help form a research question. They simply cannot carry claims that the evidence does not support. The framework does not determine employment duties, legal authority, or privacy rights. It does not authorize contact or guarantee that a correctly classified stakeholder will respond, support, approve, or buy. Role evidence is one part of a scoped buying-group map.

11. Common questions

Is "decision-maker" ever an acceptable label? Use a more precise function whenever possible: approves funds, signs contract, accepts specification, or owns the business outcome. If the exact authority is Unknown, say so. Can a title support an Inferred role? Only when combined with scope-relevant evidence, and the inference must retain its limits. Title alone should remain a clue with role Unknown. What if a person says they are the buyer? Clarify what that means in the decision: sourcing, contracting, evaluation, approval, or order administration. Record the direct statement and its scope. Do small companies need this separation? Yes, although one person may hold many functions. Separating the functions still reveals which conditions have been addressed. When does an Inferred role become Known? When direct, current, scope-matched evidence shows the person performing that function. Authority needs its own direct evidence.

12. Sources and methodology

The three-fact separation, role-evidence ladder, action table, claim fields, and falsifiable-hypothesis format are SellersGPT editorial tools. Occupational sources describe work classifications and cannot prove a named person's buying role. 1 BC05-BLS-SOC - U.S. Bureau of Labor Statistics, "How Do I Find My SOC Code?" https://www.bls.gov/soc/finding_soc_code.htm Accessed August 24, 2026. 2 BC05-ONET-CONTENT - ONET Resource Center, "The ONET Content Model." https://www.onetcenter.org/content.html Accessed August 24, 2026. 3 BC05-STATCAN-NOC - Statistics Canada, "Introduction to the National Occupational Classification 2021 Version 1.0." https://www.statcan.gc.ca/en/subjects/standard/noc/2021/introductionV1 Accessed August 24, 2026. 4 BC05-FAR-SELECTION - U.S. Federal Acquisition Regulation, "15.303 Responsibilities." https://www.acquisition.gov/far/15.303 Accessed August 24, 2026. 5 BC05-SEC-EDGAR - U.S. Securities and Exchange Commission, "Search Filings." https://www.sec.gov/search-filings Accessed August 24, 2026.