I run a PR agency that lands our clients in TechCrunch, CNBC, & other world class publications on a consistent basis.
My superpower? I can make anything sound newsworthy to the journalist I'm pitching.
These are the 2 golden rules I follow that 10x my placement rate: 👇
1/8 🧵
Most people hear “exclusive” and assume their story can only run in one place. In reality, “exclusive” means: “I solemnly swear, you get first dibs…
…you’ll be the FIRST to break this story…
…But once it’s live on the Internet, anyone else can run with it.”
And any decent PR pro will notify other publications a day or so after the exclusive story runs to squeeze out as much awareness and authority as possible.
Example:
We gave Axios Pro exclusivity on a tech client’s 8-figure funding round. Paywalled, yes… but an investor reached out the next day asking to get in on the next round.
Then I emailed Supply Chain Brain, Supply Chain Dive, Supply Chain 24/7, and more trade pubs a day later with some fresh angles that went beyond that initial story. Because Axios had already run it, they were more open to picking it up too.
In fact, landing the exclusive in a high-authority outlet first makes the broad-reach pickups easier. Other editors see that a rigorous publication already vetted the story so it lowers their risk and raises your credibility.
TL;DR
“Exclusive” in PR means “first dibs.”
Not “till death do us part.”
I just analyzed every pitch I’ve sent since Dec ‘20 (~1,225 in all). A few interesting finds, but one changed how I’ll pitch reporters from here on out.
Okay, I actually had Claude create a team of agents to do this analysis.
They read ~3,400 emails tied to every pitch I sent between Dec ‘20 and Sept ‘26. That’s the initial pitch plus all the back-and-forth after it.
Then they tagged each pitch with the trigger it used in the subject and first lines, using my MINUTES framework (Momentum, Identity, Novelty, Urgency, Tension, Exclusivity, Surprise).
Then they scored the outcome by the reporter's reply. If they asked a question, wanted more info, or outright said they’d cover it, that counted as “interested.”
Here are the most usable findings:
- Identity was the strongest trigger with cold reporters: 10.2% replied with interest. (When a reporter doesn’t know me from Adam, the fastest way in is making it about who their readers are.)
- The reporters who do know me pay more attention: 34.9% interested answers from reporters who’d previously been interested in a story compared to a 7% interested reply rate when cold.
- Pitches written for one reporter got 19.6% interested replies. The same kind of pitch sent to 8+ reporters got 3.9%. (Those wide sends were breaking-news commentary that had to go out fast. Necessary sometimes… but now I know the cost.)
And the find that changed how I pitch:
- Pitches with one clear trigger beat pitches with two: 8.5% vs. 6.8%.
(Two triggers meant two half-stories fighting over the same two lines, and the reporter had to pick. So choose your best one and cut the rest.)
That second trigger always felt like insurance. Turns out it was a tax.
Look, this is my inbox, not a controlled experiment, but from here on out, every pitch gets ONE trigger only.
I was running 7 AI agents at once when my MacBook Air started making noises I can only describe as… disturbing.
It wasn’t on fire (yet). But after a few months of hot to the touch, spinning wheel of death, Chrome refusing to open, etc., I decided it was time for an upgrade.
So I joined the club and got a Mac Mini. And I meannnnn… life-changing.
While I’m doing bedtime with the kids or chasing that mythical 8 hours of sleep for myself, my agents are drafting awards and speaking grids, building share of voice decks, and comparing last quarter’s coverage to this quarter’s.
Work that used to eat a couple of days is waiting for me as a first draft with my morning coffee.
Which means my daytime hours go to the human parts:
- Reading headlines (lots of them)
- Finding the angle nobody’s pitched yet
- Actually talking to clients and reporters
Thing is, I don’t think the real AI win is “do more.” It’s “do more of the right stuff, and close the laptop by dinner.” (Okay, close-ish. I’m still a founder.)
Let the robots work nights.
What’s the first thing you’d hand off to the night shift? Lmk in the comments.
Your competitor's Google Ads conversion values sit in a public file anyone can open. That and 8 other posts from the last 24 hours, in one place.
Every item gives you something to use today, from a Search Console regex to a 10-step blog layout.
You will not find these again by scrolling. Bookmark now.
1. @googlewaqas
Want to know what your competitor pays per demo?
Waqas looked at public Tag Manager files and found conversion names and values exposed.
https://t.co/I6EJVJVPau
2. @aakashgupta
How did one Apple decision shape American messaging?
Aakash Gupta cites court emails: Apple kept iMessage off Android to keep kids on iPhone.
https://t.co/F3BCBeYoEf
3. @akshay_pachaar
Can you cut agent costs by switching harnesses?
Akshay reports benchmarks where switching only the agent harness cut costs 25% to 64%.
https://t.co/Y6GtLcKNX9
4. @jakezward
What does a blog that converts look like?
Jake Ward shares a 10-step blog layout for SEO, AI search and signups.
https://t.co/S021DJxk4t
5. @TechSalesGuy
How do you take control of a sales conversation?
Tech Sales Guy breaks it into four moves: interrupt, compliment, common ground, then a recommendation.
https://t.co/ukISqf8NCG
6. @blvckledge
Why are some brands scaling Google Ads 3-5x faster?
Jackson Blackledge credits full-funnel search, new image formats, Demand Gen and AI Overview pages.
https://t.co/04kr2tW8yn
7. @acecreamu
Want the questions people ask before finding you?
Oleksii shares a Search Console regex that surfaces every query of eight-plus words.
https://t.co/FNKBZtxFKG
8. @PRcarly
How did Carly Martinetti build 20 Claude agents?
Carly Martinetti shares a seven-step process, with example prompts and read-only guardrails.
https://t.co/p8nPkrOwIE
9. @ENowoslawski
Which GTM tools is Eric Nowoslawski loving right now?
Eric Nowoslawski names Exa's places search, p0's Decisions API and Zenrows' new crawl feature.
https://t.co/VZQbLstggz
I do this every day: 838 posts in, 9 out. Follow me and the list lands in your feed before you go looking.
Good Morning America, CBS Mornings and the Today Show all called the same guy to explain Taylor Swift’s trademark filings. None of them found him by accident.
That guy is our client @JoshGerben, a trademark attorney.
People ask me all the time how a client ends up on three morning shows in one week. It isn’t luck, and it definitely isn’t “knowing someone at GMA.”
Here’s the process we used:
1. Pick a lane where news happens constantly.
Josh’s lane is trademark filings, which are a nonstop news machine if you know where to look. Your lane is wherever your expertise meets a recurring news moment. Find it, then own it.
2. Watch the source, not the headlines.
Josh spots filings before anyone covers them. By the time a story trends, every expert in the country is in a reporter’s inbox. Timing beats perfection every single time.
3. Pitch the commentary before the story exists.
We were the first to pitch the Swift filings with Josh’s take already attached. Hand the reporter the headline, the context, and a quotable expert in one email. The less thinking they have to do, the faster they say yes.
4. Make credibility verifiable in 10 seconds.
Producers on a breaking story don’t have time to vet you. They Google you. Josh has over 500 earned placements (Reuters, Bloomberg, CNN, BBC…), so the vetting takes about 10 seconds. Every placement makes the next booking easier.
Now, this isn’t a playbook I can guarantee for every client. Trademark law is unusually newsy. But steps 2 through 4 work in almost any industry.
I just built my 20th agent with Claude. It’s getting crazy easy. Here’s the process I follow… example prompts and all:
1. Say what you’re building and why. Keep it to one or two sentences.
Example: “We’re building a CRM that tracks every reporter I’ve pitched so I know who’s a good fit for each story and who to check in with soon just to remind them I exist.”
2. Explain how it should work. Cover the inputs, the outputs, and how you’ll use it.
Example: “Search my email and LinkedIn DMs. Track when we last talked, the stories we did, and how they like to be pitched. Make it a page where I type a name or a topic and see the history.”
3. Ask Claude to improve your plan… because it almost always does.
Example: “Rewrite this as a step-by-step plan and tell me what you’d change and what I missed.” (It added a checkpoint so I wasn't automatically paying for research on reporters who never responded.)
4. Split the work between models so it doesn’t get too spendy.
Example: “Use Fable 5.1 to handle the strategy and build, and Sonnet 5.5 for standard research.”
5. Review the plan, not the code.
Example: “Explain each step of the plan in plain English.” (Py? JS? No idea, don’t care. I check that the process matches how I’d do the work by hand.)
6. Check in when anything gets slow or expensive.
(The LinkedIn subagent spent 15 minutes opening a single thread. We decided to grab only the last DM date and a preview line. No thread summary.)
7. Make it update itself.
Example: “Set a scheduled task for Monday mornings around 6:30 to refresh the data.”
What I don’t do:
- Hand over my passwords. When it needed LinkedIn, I signed in myself.
- Let it send anything. It’s read-only on my email and LinkedIn.
- Believe what it says without proof. Every pitch preference needs a source link.
It sorted 19,411 email threads, found 3,456 journalists, identified 336 I’ve worked with, and called 30 “warm relationships” (but we all know AI’s not perfect at counting).
And it runs an update every Monday at 6:34 AM. (Close enough to 6:30, Claude, close enough.)
P.S. If you’re already building, tell me what you’d add or do differently. If you have questions, ask away. I’ll answer in the comments.
@bayareawriter Yeah, unfortunately it comes down to being able to manage client expectations (amongst other things...). I'm sorry that's happened to you so often. It shouldn't even have had to one time.
A reporter pushed an exclusive TechCrunch story due to "a personal emergency." The client asked if we should shop the story elsewhere.
I get what they were thinking. They were a brand-new VC fund, and this story had a big job to do: proof for investors during their raise, and credibility with the founders they wanted to back.
The reporter needed three weeks. And when you’re a startup, three weeks feels like three (or thirty) years.
Their worry was simple: “What if the reporter drops the ball?”
Totally fair. And the easy move was right there… quietly offer it to another outlet and hedge our bets.
But I told them that was a bad idea.
Reporters are people first. Someone had just told us they were going through something hard. Answering that by shopping their story would feel wrong, even if nobody ever found out.
In my experience, reporters are refreshingly straightforward. If they’d lost interest, they’d have told us. A three-week delay was just a delay.
We drafted a response:
“Extremely sorry and hope everything turns out OK. Please take all the time you need—we’ll be waiting.”
And to our client’s credit, they trusted us.
That’s the part of this job nobody sees. Not the pitch, but being the calm voice when everyone understandably wants to see some activity.
Three weeks later, the reporter was back. The story ran. And the client got exactly what they needed from it: credible, third-party proof to put in front of investors and founders.
PR is a long game. The wins come from the stories. The longevity comes from how you got them.