The real cost of a bad pitch isn't the rejection. It's what it does to your next pitch. Pitching often gets treated like a single transaction. You send the email. They open it or they don't. They cover it or they don't. Move on. But journalists have memories (plus inboxes with a search function). And a deep, well-earned suspicion of anyone who's wasted their time before. So here's the mental check I run before I hit send on anything borderline: - Is this pitch worth trading on my name? - Is it actually a fit, or am I just hoping it lands? - If it doesn't land, am I willing to spend this reporter on it? If the answer to any one of those feels off, I hold. Because every time you pitch a reporter something that’s not a good fit for them, you’re burning your next shot. Even if that one’s a shoe-in. Qwoted reported a 200% spike in pitch volume over the past year. Journalists are getting buried. So if a reporter knows you wasted their time already, you don’t have a shot. Think about it from the reporter's side: You finally have a pitch that's perfect for them — their exact beat, a fresh angle, just the kind of story they've been hunting for. But six months ago, you sent them something so off-base they rolled their eyes and trashed it. Why would they give your next email the benefit of the doubt? Treat your reporter list like it’s a reputation account. Every pitch is a withdrawal or a deposit. Spend accordingly.

“Press releases are back!” We've all heard someone in PR say it. They’re not totally wrong… But there's a lot more nuance than the soundbite suggests. THE CASE FOR THE COMEBACK: Muck Rack's Generative Pulse report (Dec 2025) confirmed press release citations in AI search grew 5x between July and December. PR Newswire's 2025 Global State of the Press Release Report found 93% of US communicators expect to send the same or more press releases next year. It makes some sense. Press releases are predictably formatted, keyword-dense, and timestamped. Exactly the kind of structured content LLMs love. THE PART NOBODY'S HIGHLIGHTING: That 5x increase is a jump from a whopping 0.2% to 1% of total AI citations for newswire services, and from 1.2% to 6% for all other press release citations. Journalism still drives 20-30%. And the loudest voices declaring the comeback are the ones selling press release distribution. PR Newswire launched an AI-optimized distribution product at the same time they published the report saying press releases are essential for AI. WHY THAT’S A PROBLEM: SEO operators are openly buying $80 press releases from cheap newswires, stuffing them with branded keywords, and watching them rank in ChatGPT and Perplexity. This is the exact arbitrage pattern from old-school SEO. Game the algorithm and get a temporary boost before it corrects itself. If your strategy depends on this trick, you're one model update away from being back at square one. THE SMARTER MOVE: Press releases still earn their place when you have real news. Front-load the stats. Cite specifics. Write like a human with the structure of a machine so both humans and LLMs can use it. But the kind of coverage that compounds in value is still the same: higher-authority placements, expert positioning, coverage you earn because your story has actual impact. That kind of credibility won’t get corrected-out in the next algorithm update.

82% of journalists now use AI in their work, per Muck Rack's 2026 State of Journalism report. Most PR pros think this is the story. The real story is where it’s all going wrong: ✓ AI is great at transcribing interviews. ✗ AI is terrible at pulling out quotes with strong human impact but little subject impact. ✓ AI is helping reporters research your client faster than ever. ✗ AI is also pulling outdated info from 2023 and presenting it as current fact. (The Washington Post's test of major AI tools found this was a consistent flaw across the board) ✓ AI is helping fact-check coverage at scale. ✗ AI is also hallucinating quotes and fabricating sources. (Publications from the NYT to Wired to the Chicago Sun-Times have published retractions in the last 12 months) ✓ AI can summarize a 40-page report in 30 seconds. ✗ AI also "summarized" a Guardian book review into a NYT freelancer's draft and got him fired. ✓ AI helps reporters keep up with a 200-pitch inbox. ✗ AI-powered inbox tools like Sanebox and Copilot also filter out perfectly good pitches it *thinks* are AI generated, even if they’re not. ✓ AI is changing how journalists work. ✗ AI is not changing what makes a story worth telling. ALL THIS MEANS: Your pitch is being skimmed by a human, possibly screened by a tool (one that’s confident even when it’s wrong), and potentially summarized into a draft that you'll see published with details you don’t recognize. ADVICE FOR PR PROS: → Tighten the subject line → Front-load the proof → Make the angle specific (so that an AI can't reduce it to a generic trend) The rules of engagement are changing incredibly fast. PR folks, any other tips you’d offer for our colleagues?

ICYMI: On April 30th, Bloomberg broke the story that 20 of the country's biggest news outlets (CNN, NBC, USA Today, Vox) sent a formal demand letter to a nonprofit web archive that has been quietly training nearly every major AI model on the planet. The News/Media Alliance is asking Common Crawl to remove their scraped content, prohibit AI training use, and add enforceable warnings to its opt-out registry. If you're not familiar with Common Crawl, it's the open repository of online content that Google, Meta, OpenAI, and Anthropic have all used to train their chatbots. Critics call it "data laundering." It’s the back door to gated information that publishers couldn't close. Until now. WHAT THIS MEANS FOR PR: 1. The cost of AI citations is about to go up. Anything Common Crawl scraped was effectively free training data. As publishers wall off access, AI companies will need to pay licensing fees, narrowing the pool of citable sources. Coverage in licensed pubs will still compound in value. Coverage everywhere else may not. 2. The "publish once, train forever" free ride is over. Per Muck Rack's 2026 research, half of all AI citations come from content published in the last 11 months. Pair that with publishers pulling back archives, and the half-life of an earned media hit on AI is shrinking fast. 3. Mid-market companies are uniquely exposed. Enterprise brands have massive archives and decade-old PR programs feeding the citation loop. That leaves mid-market companies competing for the remaining citation share. Every placement has to work harder, and a steady cadence of credible coverage matters more than the occasional tier-1 home run. BOTTOM LINE: When Anthropic reached a $1.5B settlement with authors last fall, I predicted that media outlets would soon be split into two tiers: those with AI licensing deals, and those without. The Common Crawl letter signals that prediction is no longer theoretical. The free-data buffet is closing. The companies building a habit of coverage in licensed publications today will be cited in AI answers tomorrow. To my PR colleagues: I’m curious, how are you factoring these changes into your 2026 strategy?

I can spot AI-generated PR work in about six seconds. I'm not even talking about em-dashes or sentence structure. I mean the smoothness with no spine. Grammatically accurate sentences that aren’t actually saying anything. Turns out there's a name for it: "workslop." In a Harvard Business Review report, Stanford and BetterUp researchers found 40% of workers are receiving some form of workslop. And it’s costing them: - About 2 hours of rework per instance - $186/month per affected worker - $9M annually for a 10K-person org This helps explain why, according to an MIT study, 95% of companies see zero ROI on AI. HOW THIS HAPPENS: An AI-generated response appears professional enough that people assume it's substantive. The sender thinks they’ve just saved hours of work. But without any real thinking behind it... the receiver of this AI dump now carries all the cognitive load of interpreting vague points, inferring missing context, deciding whether to send it back or just fix it themselves. Using AI like this doesn’t save time. It eats it alive. IT’S NOT AI’S FAULT The same research found that people with a "pilot mindset" (high agency optimism purposeful use) use AI 75% MORE than “passengers” (low agency optimism crowd). But they’re also getting BETTER results. So the gap isn't between AI users and non-users. It's between THOUGHTFUL users and... those contributing to workslop bottlenecks. WHY THIS MATTERS FOR PR PROS: Roughly half of recipients view workslop senders as less capable. 42% see them as less trustworthy. 37% as less intelligent. Almost a third say they don't want to work with the sender again. In an industry where the entire deliverable is trust… that better not be you. Instead, use AI as a thinking partner: to get a second opinion, for research (as long as you check sources), or to help polish a piece of writing. And it can absolutely boost productivity, even creativity. Just don’t outsource the substance.

@G2dotcom surveyed 1,076 B2B software buyers and found that 51% now start their research in an AI chatbot (vs. Google, up from 29% a year ago). That’s a ~76% increase from last year, and that trend is only moving in one direction. Before we unpack what this means for YOU… HERE’S WHAT ELSE THEY FOUND: - 69% chose a different vendor than they originally planned, based on what an AI chatbot told them - 33% bought from a vendor they'd never heard of before the AI surfaced it - 80% say the AI accelerated their purchasing decision WHAT THIS MEANS: We all know any kind of information gathering is increasingly delegated to AI. What’s novel here is that what buyers GET from AI is enough to override their own pre-existing preferences. My sense is that when ChatGPT recommends a vendor, the buyer isn't trusting ChatGPT itself (any more than they trust “Google”—they’re interested in what is SURFACES). They're trusting Forbes, Reuters, the WSJ, the trade publication they read every week… whatever the AI pulled from to build the answer. The AI is just the messenger. THE NEW REALITY: For a long time, brand recognition amplified earned coverage. A WSJ feature on a known brand landed harder than the same feature on a smaller company, because buyers were filtering through what they already recognized. AI pulls from credible earned media when guiding buyers along their journey IRRESPECTIVE of brand. And the answers it gives are compelling enough to drive 33% of buyers into purchasing from companies they'd never heard of. Again, I’m betting that number only moves in one direction.