Generative AI is Your New “Co-Founder.” Just Don’t Ask it to Pitch Journalists.
AI will become like a “new cofounder” for many startups, according to a recent HubSpot report.
The collaboration has already started: according to the report, the marketing function (which usually houses public relations work) uses AI most frequently and impactfully.
There’s no denying that generative AI is a game-changer for startup founders. But founders who use AI tools to quickly generate and blast email pitches to journalists won’t see any benefit.
That’s because the point of a PR program isn't to write pitches as fast (or "efficiently") as possible. It’s to raise the credibility, authority, and profile of your company.
Below, I’ll outline three main problems with AI-generated pitches, share how you can cut through the noise, and offer some alternative marketing use cases for generative AI.
Problem 1: AI Should Simplify, Not Complicate
Some people use ChatGPT to explain a complex topic to them using simple, easy-to-understand language. That’s a great way to use new generative AI tools.
But it’s dangerous to copy-paste AI-generated content and use it verbatim. Why? AI-generated content uses a lot of words, yet gives very little detail. That’s the opposite of how we write PR pitches: punchy, sub-150 word blips packed with meaning.
Remember, wordiness doesn’t correlate with expertise.
For example, say I’m working with a new client that builds VR software for the real estate and housing industry.
I prompted ChatGPT (the most popular startup AI tool, according to HubSpot) to write a pitch on that topic, and it returned a 470-word epic detailing how “VR revolutionizes the way properties are marketed, viewed, and ultimately sold.”
That’s well above my rule-of-thumb 150 words per pitch, so I (nicely) asked the machine to cut it down.
Here’s the result:
Dear [Journalist/Editor's Name],
I'm excited to share how virtual reality (VR) headsets are revolutionizing real estate. With VR, prospective buyers can immerse themselves in lifelike property tours from anywhere, eliminating travel constraints and offering a comprehensive understanding of the space. This technology enhances decision-making, increases engagement, and attracts more interest to listings. Leading agencies like [Insert Company Names] report significant boosts in lead generation and client satisfaction. For instance, [Insert Example] saw a [Insert Percentage] increase in property inquiries and a [Insert Percentage] reduction in time on market with VR integration. Virtual reality isn't just a trend; it's fundamentally reshaping how real estate transactions are conducted. I'm available to provide further insights or arrange interviews. Thank you for considering this innovative angle on the future of real estate.
Warm regards,
[Your Name]
[Your Title/Position]
[Your Contact Information]
At first glance, this doesn’t sound too bad. But a deeper look reveals it relies on empty phrases, lacks specifics, and invents examples (more on that later).
That’s the exact opposite of good PR writing. It should be punchy and specific and rely on examples instead of generalizations.
George Lawton, contributor for TechTarget, had this to say on the pitches that clog his inbox:
“Email-wise, I mostly just get a lot of long-winded blather about things I don't care about," he said. “There’s a very long wind-up and then finally, sometimes, the name and occasionally the title and LinkedIn of a source.”
Problem 2: AI Isn’t Always Truthful. That Doesn’t Mean You Get a Pass
According to the HubSpot report, ethical implication ranks among top concerns for startup founders that use AI.
That tracks on the PR front. Journalists adhere to a strict code of ethics, and they expect PR professionals to as well. If we don’t, we’re putting their careers and reputations in jeopardy.
Generative AI programs don’t feel the same way. In fact, they don’t feel at all. It’s the plot of every sci-fi movie!
Take my AI-generated VR real estate pitch.
ChatGPT just invented a bunch of stats because they sound good.
You can’t build relationships with the media that way. We don’t get to invent what stats would improve our pitch first, then fill them out later. Lawton, the TechTarget contributor, puts it well: “The main problem with GenAI tools, as they stand today, is they don't respect trust.”
John Hollon, a journalist published at Talent Management and TLNT, among other outlets, echoed the importance of trust. He said he hopes PR pros would disclose their AI usage and ensure journalists understand its role in their work.
“Full disclosure is a good thing here,” he said. “So more communication, explanation, and insight up front would help journalists ease into how AI might be used in a typical PR setting.”
Problem 3: No Personal Touch, No Response
You don’t get a second chance to make a first impression.
Say you’re a journalist with a crowded inbox. Which pitch grabs your attention?
My AI-induced “I'm excited to share how virtual reality (VR) headsets are revolutionizing real estate.”
Or, “I caught your article on RON fraud (loved the quote on proactive prevention from Ronnie Whelan) — wanted to pass on a gentle clarification regarding the second graph.”
The first intro serves no real purpose. PR people are meant to facilitate, not speculate. No one cares what we’re excited about.
The second intro marks a thoughtful connection that shows you not only paid close attention to the reporter’s work, but you have an expert source that’s ready to build on what they’ve already written.
By leaning into the genuinely human emotions that are more challenging for AI to emulate, you can craft a spicy pitch that stands out in a uniform sea of bland faux-excitement.
The upshot? AI-generated or not, the best pitches are specifically targeted to an individual journalist. In fact, targeting is one of my “three T’s” of media relations outreach>.
To make sure your pitch gets read:
- Choose your target carefully
- Get to the point quickly
- Use examples to illustrate relevance, timeliness, and expertise.
Generative AI can help with point #3.
A quick query to chat GPT for “use cases for VR software in the real estate / housing industry” unveils virtual tours, design visualization, and home staging as potential examples.
While those are the real-world examples that journalists look for in PR pitches, it doesn’t mean you can just plug and play. Be sure to proofread, massage, and adjust AI-generated content before using it.
For founders often knee-deep in technical product speak, generative AI can help simplify key concepts and relay them to the media.
Problem 4: The Writing Isn’t the Hard Part
You don’t need to be the next great American novelist (or the next great American AI model) to write a good media pitch.
If you can write at a fifth-grade level, you can write a good media pitch. But that’s the easy part.
You also need the self-awareness for what is (and isn’t) newsworthy about your organization, the interpersonal skills that help build key relationships with the media, and the understanding to pull everything together.
That’s what generative AI has such a hard time replicating: the accumulated knowledge that fuels the pitch-writing process itself.
When a successful pitch lands, it’s like when a golfer hits a perfect shot onto the green in a big tournament. To a non-golfer like me, it looks almost too easy.
But you don’t see all the bad practice shots that ended up in the bunker (or in the parking lot). You don’t see the mistakes that happen when you use the wrong club on the wrong course under the wrong weather conditions.
As a startup founder or marketer, you’ve gathered more industry knowledge than you probably realize. Further, your brain can apply that knowledge in the right context while navigating the relentless ebbs and flows of the always-on news cycle.
And with every pitch you send out (and every non-reply), you’ll develop a deeper understanding of the media landscape which you can use to hone your approach and make the most of your efforts. Continuous learning isn’t just for machines.
Thoughtful Research + Good Writing = Real Connection
LLM-backed chatbots like ChatGPT are really good at quickly producing large blocks of human-sounding, sometimes error-riddled text.
You know who else is good at making errors and sounding somewhat like an actual person? The PR professional drafting this blog.
The difference between me and a computer program, though, is that my pitches lead to relationships, as evidenced by this note from a third journalist:
“I know if I need solid expert commentary, you are the man to mine it for us.”
That’s not because I crafted the perfect ChatGPT prompt or sent him hundreds of irrelevant messages and follow-ups.
It’s because I got his attention by taking a real interest in his work, delivered thoughtful sources who gave strong quotes, and proved myself valuable to him as a reliable PR pro.
If you blast journalists with AI-generated pitches, you might get lucky and get a response that eventually leads to a placement. A broken clock is right twice a day.
But it’s harder for AI tools to help you build relationships. When journalists know they can depend on you, they come back to you for insight (often unprompted!). That’s when the “PR flywheel” (and the media ubiquity that comes with it) really starts spinning.
Journalists’ crowded inboxes didn’t start with ChatGPT. Timely, targeted pitches — and the relationships that follow them — will continue to stand out among the junk.