The digital approach to PR: Clean, recycle, repeat

Every digital PR (DPR) team has ever existed: New data is pouring in and the team is hustling while someone is staring at a blank Google doc that’s going around angles and a reporter’s target. Finally, the voice gets out the door in time to click “Send” before the end of the day.

The stage then settles to a high point, everyone is happy, and the next month the whole team does the exact same thing, like it never happened.

But here’s the thing no one talks about: That winning field is a valuable asset, and most teams will just leave it sitting in their shipping folder gathering visible dust.

Whether it was a data survey, a product launch, or an expert quote, that word is a template. And with AI, you can incorporate its DNA into every new campaign rather than reinventing the wheel every time.

In numbers

The numbers to get this right have never been higher. About 46% of journalists receive six or more comments every single day of work, and of those, 49% rarely or never respond when asked, according to Muck Rack’s State of Journalism report.

Volume continues to rise while relevance declines, with 47% of journalists saying they rarely or never find voices relevant to what they write, Cision’s 2025 State of the Media Report found.

The volume problem is real, and AI makes it worse by allowing everyone to generate pitches quickly and easily. This means that journalists’ inboxes are quickly filling up with content that sounds more familiar than ever.

So how do you get your points in front of as many journalists as possible while actually getting noticed? The answer is deceptively simple: Rather than measuring your generation blindly, measure what you already know.

Meet the DPR method of replication

I call it the “DPR method of repetition,” and the concept is simple: rinse, recycle, repeat.

The process is straightforward. You take the pitch that created the coverage before, determine what really made it work structurally, and then use AI to replicate that structure for your next campaign rather than rolling out a blank slate.

It works on all kinds of pitches, too, which is my favorite part about it. Data studies, product launches, expert quotes, practical commentary — it doesn’t matter. If a structure has worked once, it can work again, and if it has worked 10 times, it can work 20 times.

One of my favorite quotes to use in this way is one I sent to an editor at PR Daily, and the subject line reads: “Your basset hound is so cute. [New SEO study for PR Daily].”

The pitch was built around a study of data on the performance of YouTube icons, and the findings were clear, visual, and easy for the journalist to be an independent story without heavy lifting. It stayed. Same day response.

Anatomy of a successful pitch: What made it work?

So why did it work? There are four reasons, and you can repeat each one individually:

• The subject line leads with a personal connection before you speak in tone, directly targeting the editor’s dog before placing the reading hook in parentheses. This made it impossible to ignore because at first it didn’t sound like a platform. Instead, it felt like a personal message from someone who really knew them.
• The opening hook builds rapport before building a case, introducing their pet and sharing something personal before moving on to the real reason for the email. When the data came out, they were learning and accepting.
• The mathematical sequence is from broad behavioral findings down to the more specific and observable. This gave them many angles to work with, depending on what their audience needed most. It did not force them to discover the matter themselves. Moreover, it was about a topic they were already talking about.
• The CTA is built around its readers and not around my course or client. It asked if their growing business audience interested in video would benefit from the findings. The CTA wasn’t just, “Would you like to cover this?” Rather, it was, “Can your students benefit?” That’s a very different question, and journalists quickly sense the difference.

Steal the structure: Quickly quickly

Don’t dictate your best pitch to AI. Instead, give it a voice by pasting the full text. Then, ask them to point out the parts that made the pitch work rather than writing something new from scratch.

Here’s what that looks like using a guesswork campaign. It says a new survey by a financial health firm shows that one in three Americans skipped a doctor’s appointment in the past year because of the cost. This is solid data with an obvious emotional hook that most reporters covering personal finance or health care will care about.

You need to stop it, and you need it to stay down. So you open the PR Daily pitch above, and use it as your blueprint, repeating each part that made it work for the new campaign.

Repeat the subject line

That PR Daily subject line worked because you opened with something personal to the reporter before you talked about the research, and you want that same energy in every new pitch you post:

• “Create seven headlines for each given figure. For example: [paste your winning subject line format].”
• “Make this subject line the focus of it [new topic]: [paste winning subject line].”
• “Make this subject line more newsworthy based on the topics I’ve provided: [paste current subject line draft].”
• “Make this figure a headline: [paste stat].”
• “Make this article personal to the reporter [beat]: [paste headline].”

Repeat the opening hook

The opening worked because it sounded human before it sounded like a voice, and injecting that same warmth and clarity into the new campaign is as easy as showing the AI ​​exactly what you’re saying rather than trying to explain it:

• “Love this opening. Make a new opening mimic this: [paste opening from winning pitch].”
• “Here are some trending stories. Highlight this in the opening: [paste URL].”
• “Make this more open [inflation/healthcare/financially] fixed: [paste current opening].”
• “Here is another example of what is happening right now. [paste URL].”
• “Make this introduction feel like a journalist would write it and not like a press release: [paste current intro].”

Repeat the math sequence

The PR Daily pitch figures went from broad findings to very specific and surprising, giving the reporter a ready-made narrative to work with instead of a list of numbers he had to explain himself:

• “Here are my key stats: [paste stats]. Do the math to simulate this action: [paste stat section from winning pitch].”
• “Make this figure more clear and informative but not misleading: [paste stat].”
• “Rewrite these figures so that they flow like a story, starting more broadly and becoming more specific: [paste stats].”
• “Make these statistics feel more like interviews and less like a press release: [paste stats].”

Repeat the CTA

The CTA worked because it put the reporter’s readers at the center of the inquiry instead of the research or the client, and that shift in framing is something you want to include in every pitch you send:

• “Make the CTA look like this: [paste CTA from winning pitch]. The new title is [insert topic].”
• “Make this CTA more [topic] fixed: [paste current CTA].”
• “Rewrite this CTA so that it leads to what journalism students will find, not what we want to see: [paste current CTA].”
• “Make this feel less salesy and more like a genuine offer: [paste current CTA].”

Repeat follow up

Follow-ups get the exact same treatment, because there’s a version of your perfect follow-up already sitting in your sent folder. You should always use this successful trace as a model instead of writing a new one:

• “Copy this tracking and add a link [paste URL]: [paste your winning follow-up].”
• “Speak [insert trend] from [insert article] in the following: [paste follow-up].”
• “Rewrite this trace so that it leads to a new figure that we didn’t include in the first game: [paste follow-up and new stat].”
• “Make this follow-up shorter and punchier while keeping the same structure: [paste follow-up].”

Every component has a verified version already sitting in your sent folder, so use it. Reproducing the original text rather than paraphrasing it will give more reliable results, since the AI ​​won’t need to guess your voice. Instead, it has a plan.

You can repeat anything

Ask yourself what is holding your current positions back. The first answer that comes to mind is probably not the lack of a new AI tool. Instead, it’s probably a building block from something that already worked and that you stopped using when it arrived.

DPR’s recursive approach can apply to all parts of your reach (eg, headlines, voice intros, math formatting, CTAs, opt-outs, and follow-ups). Every single component can be modified from a version that has already proven its effectiveness.

I know what you might be thinking at this point: Won’t spaces start to sound the same if they all draw on the same structure? The answer is no, because the building is yours, built on your winnings, your voice, and your relationship with a particular editor about his particular dog. No one else has that plan.

Here are some questions to consider before your next campaign:

• What group of figures did you like from the previous pitch, and how can you use them as a new data formatting model?
• What pitch produced the greatest amount of printing press, and what was the reason for its operation?
• What headlines elicited responses from reporters, and what pattern did they elicit?
• What in your past experiences could be enhanced by AI rather than replaced by it?

Using AI doesn’t require giving up the secret sauce of what makes a press — because the strategy is still yours. AI just helps you use it faster and more consistently without losing some of the ingredients that made your best work work.

Your next pitch starts with your last win

Turn on the pitch that generated your best coverage in the past 12 months, whether it was a data study, a product launch, or a professional quote pitch. Identify the elements that made it work, including the subject line, opening hook, sequence of figures or story, and CTA. Notice what makes each person feel specific, human, and impossible to ignore.

Then tell the AI ​​to reproduce each part using that voice as a model. Add the current context where it fits, combine everything, improve as needed, and repeat the following.

You are not copying. He combines.

Clean, reuse, repeat.