Happy Wednesday! In today’s newsletter:

Let’s get started!

The key to getting AI chatbots to work well for you is to understand what they don’t do well. Specifically, they don’t come to the conversation with a lot of context.

It’s our job as communicators to give them that context. And I found a neat way to do it. I call it MESSAGE, and it solves the most common reason prompts fail: not enough detail, not enough direction, and not enough strategy.

Communications professionals are trained to think about audience, context, tone, reputation, narrative, and desired outcomes. But when we open ChatGPT or Copilot, we sometimes forget all that, and we ask AI to do the impossible with a seven-word prompt. The result is generic content, missing nuance, and content that doesn’t sound like you or your organization.

MESSAGE fixes that.

It’s a mnemonic device that gives you a simple mental checklist you can run through any time you need to prompt: seven components that turn a vague idea into a strong instruction. It forces clarity, protects quality, and helps you build prompts that are deeper, safer, and more aligned to how communicators think. Let’s break it down:

M — Mission
Define the exact outcome you want.

E — Environment
Provide context, background, and any relevant situational factors.

S — Stakeholders
Who is the audience? What do they care about? What do they know/don’t know?

S — Source Material
Give the model the ingredients it must use: notes, links, drafts, quotes, data, talking points, research, or prior work.

A — Angle & Persona
Tell ChatGPT what persona to adopt, e.g., “think like a speechwriter,” and define any narrative angles (optimistic, investigative, celebratory, apologetic, educational, etc.).

G — Guardrails
Protect against shallow or risky content. Include brand voice direction, language restrictions, DEI sensitivity, reading level, compliance/legal needs, accuracy expectations, format or length limits.

E — End Product
Specify what you want back, including type, length, structure, and format — e.g., a 200-word LinkedIn post or an executive summary with bullets.

Each piece represents something ChatGPT needs to know before it can produce something meaningful. If you fill in each part — even in one sentence — you suddenly move from “write something for me” to “work with me as a strategic partner.”

Here’s a simple example.

Imagine your CEO wants to announce a rebrand next week and asks you to draft the internal launch email. Normally, a prompt might look like: “Write an internal email announcing a rebrand.” But that prompt is asking the AI to read your mind — and it can’t. With MESSAGE, the prompt becomes something integrated, thoughtful, and structured:

You’re still in control. You decide the strategy. You pick the message. The model just helps you express it faster — and often, better.

And here’s the best part: MESSAGE works everywhere. Press releases, speeches, presentations, articles, CEO notes, HR announcements, social copy, crisis comms — anything that relies on words and structure. It’s discipline-agnostic by design, because comms people never know where their next assignment will land.

It also reinforces legal and reputational awareness, which is crucial in our field: you’re reminded to consider equity and inclusion and brand sensitivities before you hit “Generate.” That’s not how most prompting guides think. But it’s how communicators must think.

The big lesson here is this: prompting isn’t about typing. It’s about porting your thinking into something the AI can use. When you organize your thoughts, AI produces stronger work. When you don’t, AI fills in the blanks with clichés.

MESSAGE is simply a way to remember the thinking steps.

Your next step:

The goal is to train yourself on how to talk to the chatbots.

But wait — there’s more. I want to make this dead simple for you to use.

Download the cheat sheet ⬇

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

Download the free guide

Muck Rack released the results of a survey of 500+ comms pros, asking them how they use AI, measure its impact, and deploy it at work. You can get a copy here and my key takeaways are below.

📊 AI Adoption Has Leveled Off — But Organizational Maturity Hasn’t

AI usage among PR pros has plateaued at around 75%, but company-level support has surged, with far more organizations rolling out AI policies and training. The biggest shift isn’t who is using AI — it’s how seriously companies are formalizing it. And that means we’ll see those 75% using AI more and more.

What’s next: Expect the gap to widen between teams with clear guardrails and workflows and those still relying on ad-hoc experimentation. Competitive advantage will come from repeatable, sanctioned AI practices — not curiosity alone. (But as an individual, curiosity will still be your superpower.)

⚡ AI Is Boosting Speed and Quality — Not Replacing Human Judgment

Most PR pros say AI makes their work faster and better, especially for editing, research, writing, and strategy. Yet nearly everyone still edits AI output, even if those edits are becoming lighter over time.

What’s next: AI will increasingly handle the “first 60–70%” of comms work, while humans focus on judgment, nuance, and accountability. The strongest teams will optimize for iteration, not automation… except for when it comes to agentic AI. And on that note…

🧠 Agentic AI Remains a Trust Problem, Not a Tech Problem

Awareness of AI agents is high, but real adoption is low — and most PR pros don’t really know how to build agents, and even those who are? They’re uncomfortable letting agents act without human review. Trust hinges on human approval, transparency, and proven reliability. But that’s OK, because we’re still here to review what the machines produce!

What’s next: Agentic AI will creep into PR through tightly scoped, human-in-the-loop use cases before anything resembling autonomy. Full hands-off automation won’t land until governance and confidence catch up to capability — if ever. Having a real person review the outputs will be necessary not just for governance, but for that strategic layer critical to producing great business outcomes.

I’m low-key obsessed with generative engine optimization (GEO), or “getting ChatGPT etc. to tell your story correctly.” So when one of my newer connections, Amanda Coffee, posted about her “Mastering GEO for Comms Pros” (password: GEO) event on February 18th, I auto-clicked.

Topics include:

Featuring pros from American Express, Burson, Meltwater, Notified Investopedia, and more!

Sign up to attend in person or virtually, here — the password, again, is GEO. I’m going to hopefully attend in person, so if you do as well, say hi!

While writing this week’s newsletter, I came across Napkin, a text-to-visuals tool. I was immediately wowed and need to spend a lot more time using it, because the potential is massive. Their tagline is straightforward and on-point: “Napkin turns your text into visuals so sharing your ideas is quick and effective.”

I tried to turn the MESSAGE prompt framework into a visual to test it out. Here’s what it gave me, after some quick (very quick!) edits on my end.

👍 What worked for me: The tool was super simple to use. I gave Napkin the nuts and bolts of the MESSAGE framework, and it gave me some pretty good options, including the above. It also thought outside the box a bit and came up with this bridge-style process image, below, which I liked, mostly, but wasn’t a spot-on match for anything I was imagining (More on that below.)

👎 What, uh, didn’t: Like most gen AI tools, the output is only as good as the specificity of the input. And I don’t really love how Napkin takes inputs. I’d like a guided process that helps me refine my objectives, etc., so I can talk out what I’m looking for and ultimately get to the result I’ve either imagined or failed to yet imagine. Instead, Napkin takes your input and gives you dozens of different creative directions you can go in. That’s cool, but for me, it created a paradox of choice problem: I was so overwhelmed by the dozens of visualizations it produced, especially with no clear winner among the horde.

⚖ My verdict: I’ll use Napkin going forward, especially as it levels up. I know good design when I see it, but I’m not a designer, so this tool unlocks a new “skill” for me. I can see this becoming a go-to tool, though, as the creation process becomes smarter and more streamlined.

For now? Text-to-image visualizations are a nice-to-have. I’ll try Napkin quickly, see if I get a good initial response, and use that result; if the results are mostly a miss, I’ll just go without it.

“When I first opened Attio, I instantly got the feeling this was the next generation of CRM.”
— Margaret Shen, Head of GTM at Modal

Attio is the AI-native CRM for modern teams. With automatic enrichment, call intelligence, AI agents, flexible workflows and more, Attio works for any business and only takes minutes to set up.

Join industry leaders like Granola, Taskrabbit, Flatfile and more.

👉 Start for free today

Was this forwarded to you by a friend or colleague? Want to get this in your inbox next week? Click here to subscribe!

Until next Wednesday,

Dan