How to judge a PR agency's reporter research
By Joel Andren · Published by PressFriendly, a PR agency · Reviewed August 21, 2026 · Editorial standards · 16 of 21 in this section
Judge reporter research by whether the provider can show a current, source-linked reason that each proposed contact fits a specific story or source offer. The company approves relevance and risk limits; the provider researches and maintains the record; the journalist decides whether to respond or cover the company. List length proves little.
Cision, a PR software vendor, reports that its 2026 survey of over 1,800 journalists worldwide found 86% would reject outreach lacking relevance to their beat or audience. Treat that figure as vendor survey evidence and inspect target logic before approving outreach.
Current, specific evidence should justify every proposed contact
Ask the provider for a representative sample before work starts and before a material campaign. Each record should support these buyer checks:
| Buyer check | Evidence to inspect |
|---|---|
| Current role | Outlet page, byline, official profile, or licensed source, with the check date |
| Audience fit | Outlet product, geography, format, and the stakeholder group it serves |
| Subject fit | Recent representative work and any stated beat or contact preference |
| Offer fit | Company evidence, access, expertise, or timing that makes the contact relevant |
| Confidence | Source links, unresolved uncertainty, restrictions, and the next review date |
The provider should distinguish reporters, editors, producers, bookers, hosts, freelancers, and contributors because their authority differs. Prestige is separate from relevance. A specialist publication can serve the objective better than a large general outlet.
Provenance, privacy, and licensing are acceptance gates
Require source categories, including public newsroom material, licensed databases, direct professional contact, inference, and AI assistance. A database entry or AI summary can identify a candidate, but a human should verify identity, current work, and contact route against the source.
The contract should state which database licenses permit client outreach, what the company may view or export, who pays, and what happens at subscription or engagement end. It should prohibit sending records to unapproved AI services and define access, retention, and incident notification.
Privacy rules vary by jurisdiction, recipient, channel, and purpose. The UK Information Commissioner's Office says publicly available business-contact data can still carry data-protection duties; its guidance is under review following statutory changes. Assign a company privacy or legal owner to approve the lawful basis, notices, objection handling, retention, and cross-border use. Exclude home addresses, family details, protected traits, inferred vulnerabilities, and unrelated personal facts.
Company-specific history should survive a provider change
The provider may retain its proprietary database and methods. The company should retain a contract-defined record of approved targets, rationale, sources and check dates, outreach status, responses, corrections, commitments, and restrictions.
Name the company account owner and system of record. Require a usable exit export, subject to license and privacy limits. If a provider cannot transfer a contact field, it should still transfer commitments and enough restriction information to prevent duplicate, conflicting, or prohibited outreach. No party owns a journalist or guarantees future access.
Refresh rules should expose uncertainty before outreach
Agree on triggers for rechecking the record: a role or outlet change, a bounce, a correction or objection, a long campaign pause, a new geography, or a high-risk announcement. Records should show a last-checked date and a status such as current, stale, unverified, or restricted.
Review quality through a sample of rationale, correction speed, stale-record removals, duplicate prevention, complaints, and restriction handling. Response and coverage belong in reporting, but silence or a decline does not by itself prove poor research because the journalist controls the editorial decision. Repeated beat mismatches, inaccurate identities, and ignored preferences are stronger evidence of a research problem.
The gated Startup PR Playbook contains the practitioner workflow. Buyers can use the media landscape and pitch evaluation guidance to assess how research feeds the wider program.