River
Y CombinatorBacked by Y Combinator
FREE TEMPLATE

Media List and Pitch Template

A media list where no row exists without a dated, on-topic byline behind it, and the angle comes from what that piece concluded.

Free download  ·  No account needed

Byline Evidence

One row per journalist, created from the article

Subject: carrier settlement  ·  window: 180 days  ·  days since computed, never stored

JournalistMost recent on-topic articleDaysAngle
Calla LoxdaleBroker and CarrierQuick pay is a discount wearing a different hatArgues the fee costs a carrier more than the cash-flow gain is worth2Contrary conclusion
Quill QuarringtonMiddle Mile ReviewThe broker selection process, from the cabTreats payment terms as one factor among several4Planned
Maeve YaxleyHaulage WeeklyThe economics of paying a carrier in four daysRan the numbers and stopped short of the capacity claim15Follow-on
Wilhelm VerinderBroker and CarrierCarriers want consistent freight, not faster moneyArgues capacity is won on lane consistency, not terms53Contrary conclusion
Vasco UttleyPayments in TransitException queues are the settlement bottleneckArgues exceptions, not transfer speed, set the real clock151Planned

What the rule did to the list

240 purchased rows, tagged for the category, checked against 180 days of bylines. 81 survived. Searching the subject’s own recent coverage found 23 writers the database had never tagged. The list is those 104 rows, and nothing else goes on it.

The conclusion field is the one people skip and the one that earns the coverage, because it decides the angle. "Argues consolidation has stalled" is a conclusion. "Covers consolidation" is a tag.

Ashcombe bought 240 media contacts, every one tagged for their category, and pitched all of them. 7 replied and 2 wrote something. Before the next announcement they checked the same 240 rows against a single test: name a piece this person wrote, on this subject, in the last 180 days. 81 rows passed. 68 still write and have moved to another beat. 37 file elsewhere but nothing at that outlet. 54 have no byline anywhere. 159 rows, 66.2%, were mail going nowhere.

The fix is a change of direction rather than a better column set. Work from articles to people, because a category tag records who covered what on the day the database was compiled, and newsroom employment has fallen 26% since 2008. The article does not age that way. Bylines and dates are also on the page in a readable form: author, author.url and datePublished are recommended structured-data properties, and Google asks publishers to keep those dates accurate and not in the future.

Searching Ashcombe's subject then surfaced 23 on-topic writers the database had never tagged, so the list came to 104 rows. Every one carries what that person's last piece concluded, which decides the angle. 12 of them had recently argued against the claim Ashcombe was about to make. That group replied at 41.7% and wrote 25.0% of the time: 11.5% of the list and 37.5% of the coverage. It pairs with the press release and announcement kit and share of voice analysis.

The sheets that treat a byline as the primary key

240 purchased rows audited by outlet, the 104 that survived with their angle assignment, and what each angle produced.

List Staleness Audit

Ashcombe’s purchased database segment, every row tagged for the category, scored against 180 days of bylines. Eight of the 21 outlets shown.

OutletRowsOn-topic bylineBeat changedNone at outletNone anywhereSurvivesFound by byline
Freight Ledger19952347.4%2
The Rate Con17852247.1%3
Payments in Transit14722350.0%3
Intermodal Brief12342325.0%0
Freight Capital Weekly11522245.5%2
Cold Chain Monthly9132311.1%0
The Drayage Report8132212.5%0
The Deadhead603120.0%0
All 21 outlets2408168375433.8%23

159 of 240 rows failed, 66.2%. The four columns are kept apart because they lead to different actions. 68 rows still write, just not about this any more, and those are the rows that look alive, get mailed and never reply.

Survival rate has nothing to do with reach. The smallest outlet Ashcombe targets survived at 45.5% and one of the largest at 11.1%. Two outlets have no live coverage of the subject at all, which is a finding rather than a gap to fill with whoever is listed there.

Survivors plus each failure column must equal the purchased row count, per outlet and in total. A row that lands in two columns has been counted twice and the survival rate is wrong.

Media List

The 104 rows that survived the test or came out of it, with the two columns a bought list cannot carry. Twelve shown.

JournalistOutletRow sourceLast on-topic bylineDaysAngle
Engage the contrary conclusion  ·  12 rows  ·  11.5% of the list
Calla LoxdaleBroker and CarrierByline search only2025-10-042Contrary conclusion
Una UdallFreight LedgerPurchased list2025-09-0630Contrary conclusion
Kirra FennimoreFreight Capital WeeklyPurchased list2025-07-2375Contrary conclusion
Follow-on angle  ·  17 rows  ·  already published the planned story
Maeve YaxleyHaulage WeeklyPurchased list2025-09-2115Follow-on
Ulla ThurlowLinehaul QuarterlyPurchased list2025-09-2016Follow-on
Planned angle  ·  75 rows  ·  cover the subject, have not run this story
Quill QuarringtonMiddle Mile ReviewPurchased list2025-10-024Planned
Ilse PeverellOwner Operator LedgerPurchased list2025-10-015Planned
Pádraig TarbuckFreight LedgerByline search only2025-05-06153Planned
Vasco UttleyPayments in TransitPurchased list2025-05-08151Planned

Days since the last on-topic byline is computed, not stored. A row that was fresh in March is a different row in September, and the number has to move on its own or the list quietly rots while the spreadsheet still says it is good.

The angle comes from the article, not the person. The same journalist sits in a different group for the next announcement, which is why the assignment lives next to the byline and gets redone each campaign rather than being stored as a tier.

No row without a matching row in Byline Evidence. That is the space rule, and it is the only thing standing between this and a category tag with a name attached.

Response Log

All 104 pitched, angle assigned per person, silences logged. Read the bottom two blocks together.

AnglePitchedRepliedReply rateWroteCoverage rate
Planned angle751418.7%45.3%
Follow-on angle17529.4%15.9%
Engage the contrary conclusion12541.7%325.0%
All angles1042423.1%87.7%
By days since the last on-topic byline
0 to 30 days29931.0%413.8%
31 to 90 days411024.4%37.3%
91 to 180 days34514.7%12.9%
The same company’s previous launch, same category, bought list, one generic pitch
Previous launch24072.9%20.8%

The twelve journalists who had just argued the other way converted to coverage 4.7 times as often as the planned-angle group. They were 11.5% of the list and 37.5% of the coverage. A purchased database cannot identify that group, because the field it turns on is what somebody concluded.

43.3% of the previous launch’s sends produced four times the coverage. Reply rate 23.1% against 2.9%. At four minutes a row to find and read the last on-topic piece, 104 rows is 6.9 hours, which is 0.9 hours per piece of coverage earned.

Log the silences. A group with a high send count and no replies is a list problem; a group that engages and does not commit is a pitch problem, and counting only replies cannot tell them apart.

What's in the pack

01

Sourcing Method

Fixes the subject in a journalist's words, the byline window, what counts as a byline and what does not, before a single row gets added.

02

Byline Evidence

One row per journalist: the article, its URL, its date, the outlet, and one line on what the piece concluded. The conclusion decides the angle.

03

List Staleness Audit

Scores a bought or inherited list against the same test, splitting failures into beat changed, none at that outlet, and none anywhere.

04

Media List

The list itself, carrying days since the last on-topic byline as a computed number and the angle each person is assigned.

05

Pitch Angles

Three angles written before any mail goes out, including the per-person disagreement for anyone who has argued against the claim. The strongest angle is still a real, sourced opinion, which is what an executive's point-of-view register is built to supply instead of one invented to fit the news.

06

Outreach Guidance

The six-line pitch, what belongs in a subject line, which lines may be shared inside an angle group, and the one follow-up rule.

07

Embargo and Exclusive Terms

What an embargo has to state to be one, how an exclusive is scoped and ended, and what to do when either is broken.

08

Pitch Tracker

One row per journalist: the angle, the subject line actually used, the send date, what was offered, and the follow-up with what it added.

09

Response Log

Every outcome including the silences, whether a reply needed the follow-up, and whether a published piece was prompted by the pitch.

10

Campaign Review

Four questions in the order that stops each one overclaiming, ending in three decisions that each name the figure they should move.

How to use it

  1. 1

    Install it, or take it blank

    Hand the pack your current list inside River, or take the Word documents and CSV sheets away and work through them by hand.

  2. 2

    Audit thirty rows first

    If a list exists, sample it rather than rebuilding blind. Thirty rows produces a survival rate people believe, and it usually settles the argument.

  3. 3

    Record the conclusion, not the topic

    For every surviving row, write what the piece argued in a sentence that person would accept. A topic summary cannot decide an angle.

  4. 4

    Find the group that disagrees with you

    Read the conclusions for anyone who argued against your claim. Small group, highest yield, and the one no database can hand you.

Frequently asked questions

Is this template free?

Yes. No signup, no card, and the zip is the whole pack rather than a sample of it. Edit with AI is the other branch: send your subject and whatever list you already hold, and the agent runs the byline check against it. More packs in the template library.

What format are the downloaded files?

Ten files in one zip: five Word documents and five CSV sheets. The sheets are plain comma-separated text, so Excel, Numbers and Google Sheets all take them without an import step. Append ?format=pdf to the download link for a read-only copy if you only want to look at the shape.

How is this different from a normal media list template?

The usual template gives you a Beat column you fill in from a tag and a Notes column where a recent article is optional personalisation. Here the article is the reason the row exists. No byline, no row, and the angle is assigned from what that piece concluded rather than from outlet tier.

Does this replace a media database?

No. Use one to find candidates and verify contact details, then run every candidate through the byline test before it becomes a row. On the worked example the database was right about 81 people and wrong about 159, and it missed 23 the subject's own coverage turned up.

Four minutes a row does not scale. Why bother?

Because it scales down, which is the point. 104 rows at four minutes is 6.9 hours and produced 8 pieces, against 240 generic sends that produced 2. That is 0.9 hours per piece of coverage. A shorter list built this way costs less time in total than the long one did.

Why pitch someone who disagreed with me?

They have a public position on your subject from the last few weeks, so they are informed, currently interested and have something at stake. Ashcombe's twelve replied at 41.7% and wrote 25.0% of the time. Expect a piece that disagrees in part, and decide you can live with that first.

What does Edit with AI actually do?

It signs you up for free, drops these ten files into a private workspace, and starts by asking two things: your subject, and any list you already hold. Then it finds the recent coverage, reads the bylines and dates off it, records what each piece argued, and shows the working.

Find out how much of your list still writes about this

Download the ten files and run thirty rows through the byline test, or install the pack and let River check the list you already have.

Edit with AI