It’s 9:14 on a Tuesday. You have four Slack DMs, two calendar invites you never accepted, and an email asking if you can “just quickly” update a slide. None of this existed yesterday, and all of this feels urgent to the person who sent it, but wasn’t on your list. And you haven’t even opened your daily report yet.
Welcome to the PMM discipline. This is not a bad day. This is a day ending in Y.
Random Acts of Marketing
There’s a phrase I use that’s been kicking around the marketing world for a while, I didn’t coin it or start it, but I use it on more days than I wish: Random Acts of Marketing. And to be fair, there’s two versions of it … one to describe a marketing org without a coherent plan just throwing mud at a wall and seeing what hits, and the other to describe unplanned inbound asks. I mean the latter … although the former hits us too :)
It’s all the little (and big) asks that just pop up, quite often from some executive conversation, that suddenly land on us. Nobody involved is acting in bad faith, they’re just things that are part of working in fast moving dynamic environments, and that’s what makes it dangerous. Good intentions and a real deadline will get a wasteful side quest funded faster than almost anything else in a company.
I suspect nearly every PMM can point to things that they don’t love or are not proud of because someone with a title asked nicely (or demanded), and didn’t feel like they could say no.
The questions to always ask in these situations are:
- What does this do better for us than what we are already doing?
- What item on the agreed priority list does this bump down to the No list?
- What impacts does this have on other existing work items?
These unplanned items are not unique to PMM, they hit everyone, what makes it potentially more impactful is the overlap across Marketing, Sales and Product of PMM. This location in the middle of the Venn Diagram means we are structurally positioned to receive every request that doesn’t have an obvious owner, and the downflow impacts are felt broadly.
This position is either the best seat in the house or a landing strip for other people’s undone thinking. It depends entirely on what you do with it.
It’s an Advantage, Not a Burden
Most PMMs experience the inbound firehose as something that happens TO them. I want to reframe that, because it’s only half true.
You are the one person who sees the patterns where the same thing is asked for in multiple ways, the regularity of the asks and the aggregated and accumulated impacts. The firehose feels chaotic from inside it, but with a little distance, it’s actually highly useful data to affect change.
That’s the advantage. You’re not just the person who absorbs the asks, you’re the person with the vantage point to tell the difference between a real gap and a random act of marketing dressed up as an urgent need. Using that vantage point well means occasionally being the person who says “hold on, before we do that, what problem is this actually solving, and why is this taking over the current plan” even when that’s an uncomfortable sentence to say to someone more senior than you.
I wrote before about the “yes, but” reframe, about trading transparently instead of quietly absorbing everything that lands on your desk. This is the piece that comes before that conversation even starts. You can’t say “yes, but I’d need to drop X” if you don’t actually know what’s on your own list, in priority order, with enough conviction to defend it out loud.
You can’t trade trade-offs you haven’t tracked.
Track It or It Didn’t Happen
Here’s a hard truth: if your priorities live only in your head, you don’t have priorities. You have vibes with deadlines.
Note: this doesn’t need to be sophisticated. I’ve seen elaborate task tracking systems fall apart and a five-line spreadsheet survive, because the system that survives is the one you’ll actually update and use. What matters is that it exists somewhere outside your skull, and that it captures three things for every real commitment: what it is, who it’s for, and what it would cost to say both yes and no.
That last column is the one often skipped, and it’s the one that actually matters. When someone hits you with a new ask, you’re not deciding in a vacuum whether to jump and if so how high. You’re deciding against a visible list of everything else you already planned and committed. That’s the difference between “sure, I’ll fit it in” and “yes, but here’s what the impact is.” One is a hope. The other is a plan you can defend.
This is also, not coincidentally, how you build the case for more capacity needed. A tracked, visible backlog is the only thing that turns “I’m slammed” from a feeling into a business problem someone has to solve. I said it in the “yes, but” piece and I’ll say it again because it’s the whole game: the business responds to pain, not pleasure. An untracked backlog produces neither. It just produces a tired you, quietly overloaded, looking fine from the outside. Congratulations, you showed the business no pain, and it got its pleasure.
Where AI Fits, and Where It Doesn’t
I wrote a piece a while back about the Agentic PMM Maturity Curve, the five stages from Manual PMM grinding out every asset by hand to Source-of-truth PMM, where execution mostly runs itself and your time goes almost entirely to judgment. If you haven’t read it, the short version is that AI absorbs more and more of the execution tax as you move up the curve, and that’s supposed to buy you time back for the 20% that was always the hard part: the emotional read, the strategic call, the thing a vibe engine can simulate but never actually feel.
Here’s the twist I didn’t say in that piece. I said the volume of work goes up at every stage because AI clears the backlog fast enough that scope simply expands to fill the space. I framed that as mostly a good thing, more positioning tested, more segments covered, more launches getting the real treatment. That’s true when the expanding scope is work you chose. It’s a different story when the expanding scope is just more random acts of marketing, produced faster.
Think about what AI actually removes from the equation. It used to take real effort to say yes to a favor. Drafting something from a blank page was slow enough that the friction itself acted as a filter. You said no to more things by default, simply because everything cost you real hours. AI deletes that friction. A first draft now costs the time needed to type a prompt in your AI tool of choice, so saying yes feels free, so you say yes more often, and the backlog you thought AI would shrink quietly refills itself with lower-value work at a higher rate than before. Faster horses, same random acts.
The judgment about what’s worth doing was never the part AI could help with. It’s now the only part standing between “I move faster” and “I produce more noise, faster.” That work gets more valuable as your execution speed goes up, not less, and it’s exactly the work a tracking system, and a spine, protects.
But there’s a second, more useful thing AI can do here, and it’s not about writing your drafts faster. It’s about running your ops layer. The tracking system I just described, the list of asks, sources, and costs, is exactly the kind of unglamorous overhead that AI is good at carrying so you don’t have to. Feed it your inbound requests and let it help you sort by effort and impact. Have it flag the fourth “quick competitive question” this month so you catch the pattern before you’re four tasks deep in the same underlying gap. Let it draft the “yes, but, here’s what drops” message so the trade-off is on the table in minutes instead of getting lost in a Slack thread nobody reread. AI can’t decide what matters to your business. It can absolutely make the mechanics of tracking what matters cheap enough that you actually keep doing it past week two.
That’s the real answer to “will AI fix my prioritization problem.”
No. It will not.
It CAN carry the bookkeeping of prioritization the same way it carries the bookkeeping of execution. What’s left is the part that was always yours: the call on what’s worth your name being on.
The Role Not In The JD
Nobody hires a PMM with “professional list-tracker” in the job description. But somewhere around your second or third year in this discipline, you figure out that half the job is deciding what doesn’t happen, and doing it visibly enough that the business notices the difference between your yeses and your nos.
Three things awesome will always beat five things average. AI just changed how many things you’re tempted to say yes to before you remember that. The PMMs who move fastest through the maturity curve and still ship work worth being proud of are the ones who treat the freed-up time as a decision to be made on purpose, not a vacuum that fills itself.
That slide update from paragraph one? By Thursday it either made your defensible list or it didn’t. Either way, now you know which, and why.
What’s on your list right now that you can’t actually defend out loud? Start there.
Adam