Hiring signals · No. 08 / 12
What responsibilities should an AI search visibility hire own?
Last updated
Write the job around decisions and deliverables, not an acronym. Below is a proposed role charter for a US B2B software buyer. It deliberately distinguishes monitoring, implementation and validation.
| Responsibility | Required output | Acceptance test |
|---|---|---|
| Select commercially relevant questions | Buying-prompt coverage map informed by sales, customer research and product requirements | Branded questions do not conceal weaknesses in unbranded discovery and comparison |
| Establish the baseline | Dated answer archive and disclosed scoring method | Mentions, citations and explicit recommendations are recorded separately |
| Investigate competitor citation gaps | Exact-page source map and prioritised hypotheses | Source co-occurrence is not presented as proof of influence |
| Coordinate technical work | Engineering-ready tickets, owners and QA evidence | Recommendations progress beyond an audit deck |
| Improve product and editorial information | Approved, accurate pages addressing buying questions | Product facts, evidence and change history are traceable |
| Manage brand accuracy | Register of incorrect or unsupported claims | Each issue has a reference source, owner and recheck procedure |
| Connect observations to business evidence | Visibility, referral, conversion and customer-feedback views | The report does not equate a citation with a conversion |
| Evaluate suppliers | Comparable scopes, data-access terms and review criteria | Monitoring, implementation and independent validation are separately priced or identified |
What measurement skills should be tested in an interview?
Use a paid, bounded exercise rather than requesting a complete unpaid strategy. Give candidates a small evidence pack containing repeated answers, a flawed prompt list, a competitor citation export and a proposed intervention.
Ask them to:
- Separate a brand mention from a citation and an explicit recommendation.
- Identify missing or commercially irrelevant questions.
- Explain which results are comparable and which are not.
- Label collection failures as unknown rather than brand absence.
- Distinguish observed source patterns from causal explanations.
- Prioritise two implementable changes and identify their dependencies.
- Explain what evidence would make them stop or revise the programme.
A proposed scoring rubric is 25% measurement judgement, 25% implementation, 20% buyer understanding, 20% communication and 10% governance. These weights are an editorial recommendation, not validated predictors of job performance.
Should the hire promise special AI markup or guaranteed citations?
No such promise should be an acceptance criterion. Google’s official guidance says its existing SEO best practices remain relevant to AI Overviews and AI Mode, and that no additional eligibility requirements, special AI files or special schema are necessary. This is guidance for Google’s surfaces; it should not be generalised into claims about every assistant. (developers.google.com)
Ask candidates to distinguish documented platform requirements, useful technical hygiene and experimental hypotheses. A tactic appearing in several job descriptions does not independently validate it.
Sources cited on this page
- developers.google.com developers.google.com/search/docs/appearance/ai-features