Post 3 of 6: The AI-Augmented Product Director
There's a hidden tax on every director's calendar that nobody talks about directly: the hours you spend writing things you already know.
You know what happened in the meeting. You know the status of the project. You know what you want to ask for. But translating all of that into a well-structured email — one that's calibrated for the specific audience, hits the right tone, makes the ask clear, and doesn't create more questions than it answers — that takes time. Often more time than it should.
I've gotten a lot of that time back. Here's how.
The "Translator" Pattern
The most common place I use AI in communications is what I think of as the translator pattern. I know what I need to say. I just need help saying it in the right register for the right audience.
The move is simple:
Prompt pattern: "I need to send an email to [audience — e.g., my VP, an external partner, the engineering team]. Here's the core message: [bullet points of what you need to convey]. The tone should be [direct / diplomatic / urgent but not alarming / etc.]. Draft this email. Keep it under [X] lines."
What makes this work is the last part: being explicit about length. AI left to its own devices will write a six-paragraph email when you need four sentences. Constraining the length forces concision and almost always produces something more useful.
I then read it, adjust any wording that doesn't sound like me, and send. Total time: two minutes instead of fifteen.
Pre-Meeting Briefs
One of the biggest time sinks I used to have was preparing for cross-functional meetings — especially ones with stakeholders I don't interact with every day. I'd spend twenty minutes gathering context, another twenty figuring out what I wanted to accomplish, and then another ten trying to think through how the conversation might go.
Now I do this:
Prompt: "I have a meeting in 30 minutes with [name/role]. The purpose is [goal]. Here's the relevant background: [paste Jira ticket, email thread, previous meeting notes — whatever you have]. Give me: (1) the two or three things I need to know going in, (2) the one thing I most need to accomplish, and (3) two or three questions I should be ready to answer."
This takes me five minutes. And it almost always surfaces a question I hadn't thought to prepare for.
The third piece — questions I should be ready to answer — is the most valuable. Stakeholders almost always ask something that's predictable in retrospect. Getting ahead of it means you look prepared rather than caught off-guard.
The Status Update That Nobody Ignores
Status updates are another communication category where AI earns its keep. The problem with most status updates is that they're either too long (nobody reads them) or too vague (nobody learns anything from them). Getting that balance right requires actual craft, and most of us don't have time to apply craft to a weekly email.
My pattern:
Prompt: "Write a project status update for [project name]. Audience: [exec / cross-functional partners / team]. Here's what happened this week: [bullets]. Here's what's coming next: [bullets]. The following items need stakeholder attention or are at risk: [bullets]. Write this in 5-8 sentences. Start with the headline, not the context."
"Start with the headline, not the context" is a prompt instruction I use constantly. Left to its own devices, AI writes the way most people write status updates — with a lot of background before it gets to the point. That instruction flips it into something much more readable.
Navigating Difficult Conversations
This is the use case I don't see people talk about enough. Writing a tough email — delivering bad news, pushing back on a decision, managing an escalation — is genuinely hard. Not because of the words, but because the stakes feel high and getting the tone wrong can make a tense situation worse.
I use AI as a sounding board here, not a ghostwriter.
Prompt: "I need to tell [stakeholder] that [difficult news — e.g., we're slipping the launch date, we're descoping their feature request, there's a production issue]. I want to be direct without being alarming, take accountability without being defensive, and make a clear ask for [what I need from them]. Here's the context: [brief background]. Draft three different versions: one that's more formal, one that's more direct, and one that leads with empathy."
Getting three versions is useful because you're rarely looking for the "correct" email — you're looking for the one that fits the specific relationship and moment. Having options accelerates that judgment call.
I've used this pattern before conversations, too — when I know a difficult topic is going to come up and I want to think through how to frame it before I'm in the room.
The One Thing I Don't Outsource
I want to be clear about something: I don't let AI decide what to communicate. It helps me say things more efficiently, but the judgment calls — what to share, with whom, when, and at what level of detail — are still mine.
This distinction matters especially in healthcare tech, where the wrong communication at the wrong time can create compliance concerns, stakeholder anxiety, or misaligned expectations in ways that are hard to undo. AI can help you draft the message. It can't tell you whether the message should exist.
Also: always read what you're sending. The worst outcomes I've seen with AI-assisted communication happen when people treat it as a black box — prompt in, send out, done. Your name is on it. It should sound like you.
The Bottom Line
I estimate I've reclaimed somewhere between 30 and 45 minutes a day through AI-assisted communication — mostly on emails, prep work, and status updates. That's not a huge number in isolation. Across a week, it's meaningful. Across a quarter, it's the difference between constantly feeling behind and occasionally feeling ahead.
The real benefit isn't just the time, though. It's the consistency. The emails I send now are more calibrated, clearer, and less likely to create follow-up confusion. That's hard to put a number on, but it shows up in fewer re-explains, fewer "can you clarify?" replies, and fewer meetings that exist because an email didn't do its job.