Candidate research evidence and source quality

Candidate research is useful only when another reviewer can tell what was observed, where it was found, and how it relates to an approved requirement. A polished profile summary can conceal weak provenance. This brief therefore treats research quality as an evidence problem, not as a volume target.

Three source-led conclusions

First, the role must define the search. The US Office of Personnel Management describes job analysis as the foundation for identifying duties and competencies. O*NET and the Bureau of Labor Statistics provide occupational context, but neither source can establish that a particular profile satisfies an employer's requirement. Their classifications are reference points, not candidate assessments.

Second, public availability is not the same as unrestricted usefulness. The NIST Privacy Framework asks organizations to identify and manage privacy risk, while the Federal Trade Commission advises businesses to collect only information they need and to dispose of it securely when it is no longer required. A researcher should not copy family details, photographs, inferred age, or unrelated activity merely because a page exposes them.

Third, a sourced statement can still be unsuitable for a hiring decision. EEOC guidance and the Uniform Guidelines focus attention on job related selection practices. A citation proves provenance, not relevance, accuracy, or legal permissibility.

What a defensible research note looks like

Use a small claim ledger instead of a free-form biography.

For example, a public portfolio may directly show that a person authored a named project page. It may not establish how much of the work they performed, whether the project is current, or whether they want a new role. The note should preserve those distinctions. If two sources conflict, record the conflict rather than choosing the more convenient account.

Review the evidence, not the person

A practical quality review can begin with several research records from the same requisition. Remove the researcher's conclusions and ask a second reviewer to reconstruct them from the cited pages. The reviewer should test four questions:

1. Can the exact supporting passage or artifact be located?
2. Was it observable on the recorded date?
3. Does the role file explain why the information matters?
4. Does the note avoid unsupported conclusions and irrelevant personal data?

Failure on one question should produce a process correction, not a negative candidate rating. A dead link may require reverification. An ambiguous role criterion belongs with the hiring owner. A privacy concern belongs with the authorized privacy or legal owner. Researchers should not silently repair missing evidence by searching for more personal information.

Source quality is contextual

Government occupational descriptions are generally stronger for describing an occupation than for proving individual experience. An individual's portfolio is direct evidence of what the page displays, but may be self-reported. A third-party directory may be useful for discovery while being weak support for a decisive claim. Assigning a single universal credibility score would hide those differences.

The CIPD talent and recruitment factsheets also place sourcing inside a broader workforce and recruitment process. That context matters: candidate research should support an approved process rather than create an unofficial parallel selection system.

Governance choices for remote teams

Remote sourcing can involve researchers, recruiters, and hiring managers in different locations. Give each role a narrow responsibility. Researchers capture relevant observations and provenance. Recruiters decide whether the record meets the approved sourcing threshold. Hiring managers own role requirements. Privacy, accessibility, or legal specialists handle issues within their authority.

Version the role criteria and the research template together. If a requirement changes, do not quietly reinterpret old records. Mark which version governed each note and decide whether a fresh review is necessary. Restrict access to people who need the record for the stated hiring purpose, and apply the organization's retention rule rather than keeping a permanent dossier.

Limitations

This review does not establish that online research improves hiring outcomes. Public profiles are incomplete, self-selected, and often stale. Some people have little public presence for reasons unrelated to capability. Search engines and platforms expose different results by geography, account, and time. The cited US guidance does not resolve every jurisdiction's privacy, employment, or records rules. Employers should obtain advice for their circumstances.

The controls above assess traceability and bounded use. They do not validate a selection procedure, prove a candidate claim, or justify inferring protected characteristics. No yield, accuracy, or time-saving estimate is offered.

Method and dated source notes

The editorial team reviewed the following public materials on 2026-10-13. Publisher dates or versions are included when identified. We used occupational sources for role context, selection guidance for job relevance, and privacy guidance for collection and retention boundaries. We did not combine unlike sources into a numerical benchmark.

1. BLS, Human Resources Specialists (2025 edition)
2. O*NET, Human Resources Specialists (updated 2025)
3. OPM, Job Analysis (accessed 2026-10-13)
4. OPM, Assessment and Selection (accessed 2026-10-13)
5. CIPD, Talent Factsheet (2024)
6. CIPD, Recruitment Overview (2024)
7. EEOC, Employment Tests and Selection Procedures (2007)
8. Uniform Guidelines on Employee Selection Procedures (current eCFR, accessed 2026-10-13)
9. NIST Privacy Framework (version 1.0, 2020)
10. FTC, Protecting Personal Information (2016)

Frequently asked questions

Is a public profile automatically an acceptable hiring source?

No. Public visibility does not answer whether collection is necessary, accurate, job related, permitted, or appropriately retained. Apply the approved purpose and applicable rules.

What should happen when a claim cannot be verified?

Label it unverified and keep it out of decisive reasoning unless an authorized reviewer obtains suitable evidence. Absence of online evidence is not evidence that a candidate lacks a capability.

Can researchers summarize several pages into one conclusion?

They can provide a clearly marked synthesis, but the underlying observations and links should remain visible. The hiring owner should be able to distinguish direct facts from the researcher's interpretation.

If your team needs help defining a bounded research workflow, book a free recruiting call. Hiring and employment decisions remain with your authorized leaders.