Media targeting guide
How to build a media list for PR: a step-by-step AI targeting workflow
A strong media list is not a directory export. It is a campaign-specific hypothesis: these journalists are likely to care about this story, and the agency can show the evidence behind every important target.
I would rather see a PR team work from 35 defensible targets than export 350 names and hope volume compensates for weak relevance. A good media list should feel like an argument: we know why each priority journalist belongs here, and we can show the evidence.
AI-assisted media list building uses the agency's existing contacts as a starting point, validates whether those records are current, and expands the list using recent editorial evidence when the brief needs new relationships.
Key takeaways
- Start with owned relationship data.
- Separate contact validity from story relevance.
- Require evidence for every new AI-discovered target.
- Use recent coverage and competitor stories to find adjacent journalists.
- Review the list as a portfolio of reasons-to-care, not a directory export.
Why bigger media lists often perform worse
Large lists create the illusion of coverage while increasing irrelevant outreach, follow-up burden and sender risk. A tighter list makes personalization easier and gives the team a clearer reason for each target.
Search your own database first
Agency-owned contacts carry relationship context an external directory cannot reproduce. Search that asset semantically before buying or discovering more names.
- Past replies and relationship notes
- Current beat and affiliation
- Previous client or competitor coverage
- Geography and format fit
Use evidence to expand the universe
When the owned database is thin, compare the campaign brief with recent articles, recurring themes and competitor coverage. The recommendation should show the evidence behind the match so a practitioner can accept or reject it quickly.
What a reviewable AI recommendation looks like
A useful recommendation includes the journalist, outlet, current role, recent evidence, relevance explanation and confidence—not just a score. That makes AI a research accelerator rather than a black-box list generator.
Best practice: define the targeting thesis before searching names
Start with the story rather than the database. Write down the core news, audience, geography, subject, proof points and likely editorial formats. A funding announcement for a fintech infrastructure company needs a different target universe from a consumer-money trend using the same company as an expert source.
This targeting thesis becomes the filter for every candidate. It also makes AI discovery more precise because the model is matching against an explicit editorial need rather than a vague industry keyword.
- What is the story in one sentence?
- Which audience would genuinely care?
- Which beats and adjacent beats cover this problem?
- What geography and outlet format matter?
- What type of journalist could use the evidence: news reporter, analyst, feature writer, newsletter author, podcaster or creator?
Step 1: search the agency's relationship history
Begin with people the agency already knows. Prior replies, meetings, coverage and client history provide context that a third-party database cannot recreate. Use semantic search across notes and campaign history so useful relationships are not missed because an old record uses different category labels.
Do not automatically rank a familiar contact above a relevant new one. Relationship is one signal; current editorial fit is another. The list should balance both.
Step 2: find journalists from recent editorial evidence
Look for recent articles, newsletters, podcasts and competitor coverage that intersect with the campaign concept. Recency matters more in fast-moving beats, while recurring subject expertise may matter more in slower specialist areas.
For every discovered journalist, save the evidence that justified the match. That evidence can later support personalization and helps the practitioner reject weak AI recommendations quickly.
Step 3: score relevance with multiple signals
Avoid one opaque relevance score. Review the dimensions that create the score: topic fit, recency, format, geography, outlet role, relationship context and whether the journalist is actively covering this type of story. A 90/100 score is less useful than a clear explanation of why the person belongs on the list.
Use negative signals too. A journalist may cover the broad category but focus on a different audience, have moved beats, recently rejected the same angle or be suppressed for that client.
Step 4: validate contactability separately from relevance
Once the editorial shortlist is credible, verify the operational details. Check current affiliation, contact channel, email validity and suppression state before sending. Do not let a valid email become a proxy for relevance; validation answers whether a message can be delivered, not whether it should be sent.
Step 5: segment the list by reason to care
Segmentation should change the pitch, not just the spreadsheet color. Group journalists by the editorial reason the story fits: breaking-news relevance, data angle, executive expertise, local consequence, product review, trend context or another defensible reason.
This lets the agency prepare a small number of strong message variants rather than mechanically generating a unique but superficial email for every person.
Common media-list mistakes
The most common mistake is optimizing for list size because it is easy to count. Other failures include stale beat labels, duplicate contacts, generic personalization, treating outlet prestige as a substitute for relevance and keeping no record of why a target was selected.
A smaller list can be strategically stronger if each target has current evidence and the agency can learn from the outcome.
The PRmatic approach to media list building
PRmatic is designed to build targeting from three layers: the agency's owned media database, its client-specific relationship history and on-demand semantic discovery from current editorial evidence. The system should show the practitioner why a contact is suggested and preserve that evidence with the campaign.
After outreach, replies and coverage are intended to update the relationship context so the next campaign starts with more knowledge than the last. This turns list building from a repeated search task into a compounding agency asset.
- Search owned contacts before external discovery.
- Use recent editorial and competitor evidence to expand the list.
- Show match rationale and provenance, not only a score.
- Combine editorial fit with client-specific relationship context.
- Validate contacts close to send time.
- Feed replies, follow-ups and coverage back into the CRM timeline.
Sources & further reading
- ContentGrip — How to build a media list journalists actually want to hear from — Recent guidance on relevance-led media list building.
Maya Bennett · PRmatic Editor. Maya Bennett is PRmatic’s editorial persona: a consistent practitioner-style voice for research-backed guides produced from public sources, product research and PRmatic’s approved product capabilities.
Related PR guides
PRMATIC FIELD NOTES
Get practical PR + AI field notes.
Occasional, useful guidance on media intelligence, AI workflow, outreach and measurement. No product-access commitment required.
FAQ
Frequently asked questions
Should an AI-generated media list be sent automatically?
No. Discovery and sending are different decisions. A PR professional should review relevance and relationship context before outreach.
How recent should journalist evidence be?
It depends on the beat, but more recent evidence should generally carry more weight when a topic changes quickly.
How many journalists should be on a PR media list?
There is no ideal number. The list should be as large as the set of journalists for whom the agency can defend current editorial relevance. Quality and reason-to-care matter more than raw volume.
What should a media list include besides name and email?
Include outlet and role, recent relevant coverage, beat context, geography, relationship history, relevance rationale, contact-validation state and suppression status.
PRMATIC.AI
Not another generic AI writing assistant.
PRmatic connects media data, discovery, pitch preparation, outreach and coverage so the context behind a campaign carries forward instead of being rebuilt at every handoff.