Here is a number that should reset how you think about media outreach. In 2025, journalists in the UK and US lost at least 3,434 jobs across newsroom layoffs tracked by Press Gazette. The pace in 2026 is already faster. Two-thirds of the way through Q1, the count was running ahead of the prior year, with cuts at the Washington Post, Vox, CNBC, the Wall Street Journal, Politico and Nexstar — the kind of mastheads digital PR teams pitch every week.

If your journalist database was last refreshed nine months ago, somewhere between 12% and 25% of those contacts have moved roles, beats or publications. Some are no longer in journalism at all. And the email addresses that used to work? Many of them now bounce, hit spam traps, or — worse — go to a successor who flags your pitch as a cold lead they never asked for.

This guide is not another listicle of “top 10 media databases.” It is a working playbook for building, maintaining, and using a journalist database that actually produces coverage in 2026 — with real costs, real workflow benchmarks, and the specific failure points that kill most digital PR campaigns before the first pitch goes out.

Why most journalist databases are broken in 2026

Three structural shifts have made the traditional approach to media databases significantly less effective than it was even two years ago.

1. Newsroom contraction is permanent, not cyclical

Press Gazette’s tracker confirms 3,434 journalism job cuts in 2025, on top of 3,875 in 2024 and roughly 6,000 in 2023. That is more than 13,000 newsroom positions cut across the UK and US in three years. Bureau of Labor Statistics data quoted by Nieman Reports puts cumulative losses at more than 1 in 10 editors and reporters since 2022.

The implication for digital PR is straightforward. Static databases — even premium ones — degrade faster than ever. A media list built in January is meaningfully less accurate by June, and substantially less accurate by year-end. If your team treats a database as a buy-once-and-pitch asset, you are pitching ghosts.

2. AI Overviews have changed the journalist’s calculus

Google’s AI Overviews now appear on a majority of informational queries, and click-through rates to news sites have measurably declined. Journalists, increasingly aware that fewer readers click through, are more selective about what they cover. Press Whizz data suggests editorial rejection rates are up 33% since 2023, partly driven by AI content saturation and editor scepticism toward AI-written pitches.

What this means in practice: the bar for what gets a journalist’s attention is higher than it was. A database that helps you spray pitches to 200 contacts is now actively counterproductive. A database that helps you find the 12 specific journalists who have written about your exact angle in the last 90 days is worth significantly more than it was in 2022.

3. Database accuracy varies more than vendors will admit

Cision claims 1.2 million journalist contacts. Buzzstream’s review of media databases pegs Cision at roughly 850,000 active contacts. Capterra reviews repeatedly note that significant portions of any large database go stale within 6–12 months. Muck Rack, which has 300,000–600,000 contacts depending on the source, is generally rated as more accurate than Cision but smaller. Roxhill, the UK-focused option, holds around 190,000 manually-verified UK profiles.

The honest picture: there is a clear trade-off between database size and database accuracy, and no vendor sits on both axes simultaneously. Choosing a database without understanding which trade-off you are making is the single most expensive mistake in digital PR procurement.

Build vs. buy: the 2026 economics

The buy-vs-build question has shifted in 2026. Here is what each path actually costs, with current pricing where it is publicly disclosed.

Buy: enterprise media database pricing

Platform	Annual cost (USD)	Database size	Best for
Cision	$7,200 – $25,000+	~850k – 1.2M	Global enterprise PR teams
Muck Rack	$10,000 – $15,000	~300k	Quality-first US/global teams
Meltwater	$15,000 – $20,000	~270k news sources	Monitoring + database combined
Roxhill	£6,000+ (~$7,500)	~190k (UK-focused)	UK digital PR agencies
Prowly	$258/mo (~$3,096/yr)	~1M (variable accuracy)	Small/mid teams, no annual lock-in
Anewstip	$150/mo (~$1,800/yr)	Mid-tier	Solo practitioners, startups
Propel PRM (Essentials)	$2,400/yr	Mid-tier + CRM	Single-user with CRM workflow

Two things to note from this table. First, the gap between the cheapest and most expensive options is roughly 14x — and the database accuracy difference is not 14x. Second, none of the enterprise options publish pricing publicly. Vendr’s transaction data shows that Muck Rack discounts are routinely available when buyers actively evaluate Cision and Meltwater in parallel, with onboarding fees of $1,000–$5,000 typically on top of the headline subscription.

Build: the manual approach (and what it actually costs)

The build path uses Twitter/X, LinkedIn, journalist bylines, and email finder tools to construct a custom database from scratch. The economics are different but not always cheaper than people assume.

A realistic cost breakdown for a 500-contact custom database in 2026:

Twitter/X Premium for advanced search: $8/mo
LinkedIn Sales Navigator (essential for journalist beat tracking): $99/mo
Hunter.io or similar email finder (paid tier for 500–1,000 lookups/mo): $49–$149/mo. Full breakdown of email finding workflow in our guide to finding anyone’s email address.
VA or junior researcher time (8–12 hours per 100 contacts at £15–25/hr): £120–300 per 100 contacts
Rolling verification: 4–6 hours/month at £15–25/hr = £60–150/mo

Total first-year cost for a 500-contact custom database, including build and maintenance: approximately £1,800–4,500. That is materially less than Cision or Muck Rack, and the data is fresher because you built it for a specific purpose.

The catch: a custom database covers a narrow vertical well and a broad one badly. If you are pitching across consumer tech, fintech, lifestyle and B2B SaaS in the same week, a 500-contact custom database will run out of relevant journalists fast. If you are pitching one defined niche (UK personal finance, US enterprise SaaS, UK small business publications), a custom database will outperform any enterprise tool on relevance and reply rate.

The 2026 framework: a four-tier journalist database

Most teams treat a database as a single flat list. The teams winning in 2026 structure it in four tiers, each with different criteria, contact frequency, and pitch type. This is the structure to build, regardless of whether the underlying data sits in Muck Rack, Airtable, or a self-hosted Notion.

Tier 1: Inner circle (10–25 journalists)

Journalists you have a direct relationship with. They have published at least one of your stories in the last 18 months, replied to your last three pitches, or you have met them at an event.

Database fields for Tier 1:

Direct email + phone (where shared)
Personal Twitter/X handle and last 5 active posts
LinkedIn URL with last role change date
Beat (current, not last year’s)
Pitch history (yours): dates, subject lines, outcomes, reply timestamps
Editorial preferences: pitch by 10am UK, prefers data-led angles, hates exclusives that pre-leak, etc.
Personal context: covered the same beat at previous publication, has stated views on a topic

Tier 2: Active beat coverage (75–150 journalists)

Journalists who have written about your topic in the last 90 days. They are warm, not cold — but you do not have an established relationship. The 90-day window is the key constraint here. Beyond 90 days, the relevance signal degrades fast in 2026’s compressed news cycle.

Build Tier 2 by running these searches monthly:

Google News for your topic + last 90 days, filtering by domain authority and excluding aggregators
Twitter/X advanced search for journalists posting about your topic
Muck Rack or equivalent: filter by beat keyword + recent activity
Industry-specific verticals: ResponseSource (UK), Qwoted, Help A B2B Writer for incoming requests

Tier 3: Vertical authority (200–400 journalists)

Journalists at publications that consistently cover your space, but who have not necessarily written about your specific topic recently. Tier 3 is the bench — the people you pitch when you have a major story rather than a routine angle.

Tier 3 is where most enterprise databases earn their cost. Building 300 vertical-authority contacts manually is a 60–80 hour project. Pulling them from Cision or Muck Rack with the right filters is a 2–4 hour project. The trade-off is accuracy — expect to verify 30–40% of Tier 3 contacts before pitching.

Tier 4: Wildcards (50–100 journalists)

Journalists outside your normal vertical who occasionally cover your topic from a different angle. A consumer tech reporter who has written one piece on B2B SaaS pricing. A finance journalist who covered a tangential trend story. Tier 4 is for opportunistic pitches when you have a story that crosses verticals — and these contacts often produce the highest-DA placements because they are pitching their editor on a story idea, not a category.

The verification system: what to check before every pitch

Industry data on email deliverability is sobering. Cision’s database is widely reported to have 20–30% bounce rates on mid-tier journalist queries. Muck Rack performs better at 8–12%, but those numbers are still high enough to damage sender reputation if you pitch in volume. Roxhill, with manual UK verification, runs 5–8% — the best in the industry, but only for UK contacts.

The fix is not picking a better database. It is verifying every contact before each pitch. Here is the five-point check that takes 90 seconds per contact and brings bounce rates under 5% regardless of which database you start from.

#	Check	How	Time
1	LinkedIn current role	Confirm publication + start date matches your record	20s
2	Recent byline check	Last published article on the masthead in last 30 days	15s
3	Beat alignment	Last 5 bylines match the angle you are about to pitch	20s
4	Email syntax & domain	Run through Hunter Verifier or NeverBounce for SMTP check	10s
5	Twitter/X recent activity	Active in last 14 days; no public “please don’t pitch” notice	25s

Done in this order, the check takes about 90 seconds and prevents the four most common failure modes: pitching someone who left the publication, pitching someone who has stopped covering the beat, pitching a dead email address, and pitching someone who has explicitly said they are not taking pitches.

The end-to-end workflow that produces coverage

Database structure is necessary but not sufficient. The teams getting placement-rate ratios above industry average are the ones with a defined sequence from story to coverage. Here is the workflow, broken into the seven stages that actually matter.

Stage 1: Story selection

Before touching the database, the story must pass three tests. Is there original data, an unusual angle, or a contrarian finding? Is the angle defensible if a journalist asks for sources? Does it tie to a current news cycle or an evergreen reader question? If two of these three are missing, the story is a press release, not a digital PR campaign — and a journalist database will not save it.

“Industry benchmark: digital PR campaigns with original data earn 3–5x the placements of campaigns built on commentary alone, according to Editorial.link’s 2026 survey of 518 SEO professionals.”

Stage 2: Tier-aware list selection

Match the story to the right tier. Inner circle (Tier 1) gets the exclusive or first-look. Active beat (Tier 2) gets the broader pitch on day one. Vertical authority (Tier 3) gets a follow-up wave on day three or four if Tier 2 has produced strong coverage. Wildcards (Tier 4) only get pitched if the story has cross-vertical legs.

This sequencing matters because journalists in Tier 1 and Tier 2 read each other’s coverage. If you pitch all four tiers simultaneously, you signal to Tier 1 that they are not getting an exclusive — and exclusives are still the single highest-converting pitch type in 2026, with reply rates 4–6x higher than broad pitches per Aira’s State of Link Building.

Stage 3: Pitch personalisation

Personalisation in 2026 is not “Hi {{first_name}}.” It is a one-line reference to something specific the journalist has written or posted in the last 30 days, plus a one-line connection to your story. Three sentences total before the pitch itself. The full template library and reply-rate data is in our companion guide on cold email outreach for SEO.

Stage 4: Send timing

Press Whizz data shows journalist email open rates are highest between 8:30am and 10:30am in the journalist’s local time zone, with a secondary peak at 1:30pm–2:30pm. Friday afternoon and Monday morning are the worst windows — pitches sent Friday after 3pm have 31% lower open rates than the weekly average.

If your database does not have time-zone data per contact, this is the highest-leverage field to add. A pitch sent at 9am UK time to a US journalist arrives at 4am their time and lands at the bottom of an inbox that gets 200 emails before they wake up.

Stage 5: Follow-up sequence

Hunter.io’s 2025 outreach data shows that two-step sequences (one initial pitch + one follow-up) generate 65% more replies than single sends. Three-step sequences add a further 22%. Beyond three sends, additional follow-ups produce diminishing returns and start to damage sender reputation.

Recommended follow-up cadence for journalist outreach:

Day 0: initial pitch
Day 3: short follow-up — “Adding two data points you might find useful”
Day 7: final follow-up — “Releasing this to the wider list Friday, wanted to give you first look”

Stage 6: Tracking and database update

Every send updates the database. Reply timestamps, coverage links, requested follow-up info, and explicit “don’t contact again” flags must go back into the contact record within 48 hours. Teams that do this consistently see Tier 1 conversion rates climb from ~15% in year one to ~35% in year three. Teams that don’t see Tier 1 quality degrade as the original relationship-builders leave the company.

Stage 7: Quarterly database hygiene

Once per quarter, run a full hygiene pass. Re-verify all Tier 1 and Tier 2 contacts against LinkedIn. Drop any Tier 3 contact who has not published a relevant byline in 180 days. Pull new Tier 2 candidates from the last 90 days of beat coverage. Most teams skip this and watch reply rates decay 8–12% per quarter.

Beyond the database: 2026’s high-yield supplementary sources

A journalist database is the foundation, but in 2026 it is no longer the only or even the primary source of journalist contact. The four supplementary sources below are what separate databases that produce 5 placements a month from databases that produce 25.

1. Reactive sourcing platforms

Connectively (the platform formerly known as HARO), Qwoted, ResponseSource, and Help A B2B Writer all let journalists post live requests for sources. Reply rates are dramatically higher than cold pitches because the journalist is actively looking. The trade-off is volume — these platforms produce 10–20 viable opportunities per week per vertical, no more. The full mechanics, pricing, and current effectiveness data are in our guide to using HARO in 2026.

2. Twitter/X journalist requests

The hashtags #JournoRequest, #PRRequest, and #URGENT continue to produce viable leads in 2026, particularly for UK lifestyle, finance, and tech beats. Indian and South Asian journalists have shifted heavily to X for source requests as well. Set up a saved column in TweetDeck or X Pro for these hashtags and check 3x daily.

3. LinkedIn beat tracking

Sales Navigator with a saved search for “editor” or “reporter” + your industry keyword + recently posted in last 30 days is the most underused journalist discovery method in 2026. The best 50 contacts you build this way will outperform the 500 you import from any database, because LinkedIn surfaces journalists who are actively engaging on your topic right now.

Substack writers with audiences over 5,000 are functionally journalists in 2026 — they have editorial standards, they cover beats, and they earn links to your site that count fully in Google’s PageRank model. Most journalist databases under-index on Substack writers. Building this segment manually using Substack’s leaderboards by category produces 30–60 high-quality contacts per vertical that no enterprise database currently lists.

Five common database failures and how to fix them

Failure 1: Over-reliance on a single platform

Teams that source 80%+ of their contacts from one database hit the same journalists their competitors hit, with the same templates, in the same week. Diversification across two databases plus manual sourcing increases reply rate 18–25% on average.

Failure 2: Never updating beat data

The beat field is the most-stale field in any journalist database. Journalists shift coverage every 12–18 months. A reporter who covered fintech in 2024 may now cover crypto, or AI, or have moved to a different masthead entirely. Treat beat as a 90-day field that requires quarterly verification.

Failure 3: No time-zone or geography field

Every contact needs a time-zone tag. UK journalists pitched at 4am their time will not reply. US journalists pitched at 7pm their time will see your email after 200 others. This is the single highest-leverage field you can add to a database.

Failure 4: Treating database and CRM as separate

Every pitch outcome — reply, ignore, coverage, decline — must update the contact record. Teams that run pitches out of Pitchbox, BuzzStream or Mailshake but never sync results back to the master database lose 40–60% of the data value of their database within 18 months.

Failure 5: No “do not contact” enforcement

When a journalist asks to be removed, removing them from the active list is necessary but not sufficient. The contact must be flagged so they can never be re-imported in a future database refresh. Teams without this discipline get blacklisted by editors who notice the same brand pitching them after they explicitly opted out.

The 30-day build plan

If you are starting from scratch, here is a tested 30-day sequence to build a working journalist database from zero to first placement.

Days 1–3: Define vertical, sub-niches, and target publication tier list
Days 4–7: Audit existing data — old spreadsheets, CRM exports, past coverage records
Days 8–14: Build Tier 1 (10–25 contacts) using LinkedIn + Twitter + past coverage
Days 15–21: Build Tier 2 (75–150 contacts) using 90-day beat search + email finder workflow
Days 22–25: Build Tier 3 (200–400 contacts) — either via database trial or manual research
Days 26–28: Verify all Tier 1 and Tier 2 contacts using the 5-point check
Day 29: Run first pitch wave to Tier 1 only — 10–15 contacts, exclusive angle
Day 30: Measure reply rate and coverage; iterate before scaling to Tier 2

Realistic outcome of a 30-day build for a focused vertical: 250–500 verified contacts, first 1–3 placements from Tier 1, and a working baseline for the next quarter’s outreach.

FAQs

How much should I budget for a journalist database in 2026?

For solo practitioners and startups, $1,800–3,600/year using Anewstip or Prowly plus manual research is realistic. For agencies running 3–5 active digital PR campaigns, $7,200–15,000/year for Muck Rack or Cision plus internal verification time. For enterprise PR teams running 10+ campaigns, $20,000+/year is normal — but the cost of a single VA dedicated to data hygiene (~$30,000/year) often outperforms a marginal database upgrade.

How accurate are the major journalist databases really?

Realistic accuracy as of 2026: Roxhill at ~92% (UK only), Muck Rack at ~88%, Cision at ~75–80%, Meltwater at ~70%. “Accuracy” here means the contact is currently at the listed publication, currently covering the listed beat, and the email is currently active. Anything below 90% accuracy means you must run the 5-point verification check before each pitch — there is no shortcut.

Is a paid database worth it for a small in-house team?

It depends on scope. If you pitch one defined vertical (UK personal finance, US enterprise SaaS), a 500-contact custom-built database for ~£2,500/year will outperform any enterprise tool. If you pitch across 4+ verticals, the breadth of an enterprise tool is genuinely necessary. The break-even point is around 3 active verticals.

What’s the single most important field in a journalist database?

Last-published-date on the relevant beat. Beat alignment from 2024 means nothing in 2026. Whether a journalist has published a relevant story in the last 30 days is the strongest single predictor of whether they will reply to your pitch.

How often should the database be refreshed?

Tier 1: weekly LinkedIn check, monthly full refresh. Tier 2: monthly beat verification, quarterly full refresh. Tier 3: quarterly. Tier 4: every 6 months. Teams that follow this cadence maintain reply rates within 5% of their baseline; teams that refresh less often see 15–25% reply-rate decay year-on-year.

Should I include broadcast and podcast journalists?

Yes — but in a separate tier. Broadcast and podcast contacts have different lead times (4–8 weeks vs 24–72 hours for digital), different pitch structures, and different success metrics. Treat them as a parallel database, not a sub-segment of your digital list.

Closing read

The journalist database is having its second decade — and it is changing more in 2026 than at any point since the launch of HARO in 2008. Newsroom contraction, AI Overviews, and rising editorial scepticism mean the team that wins is not the one with the biggest list. It is the team with the freshest, most-segmented, most-verified list, used inside a workflow that respects journalists’ time and produces coverage they can defend to their editor.

Build the four-tier structure. Run the 5-point check. Layer in reactive sources. Update the database after every pitch. Do this consistently for two quarters and reply rates will outperform the industry baseline by a factor of 2–3x. For the broader strategic context, see our parent guide on digital PR for link building, and for the tooling stack that supports the workflow described here, our review of the best link building tools in 2026.