You get to the room 5 minutes early…because you always get there 5 minutes early. The lead chair is already taken by a laptop, its owner still in the hallway laughing at something you can’t hear. You take the chair with the broken tilt lock and will spend the rest of the meeting fighting to maintain some semblance of good posture—a metaphor for what’s to come.
Sarah pulls open the roadmap. “So we’re thinking Q3 for the migration,” she says, and you notice the we—how it includes you the way a photograph includes its background.
It’s already been decided, you can tell by the confidence of the performance: no questioning inflections of the voice, no pauses, slides not even on screen long enough to read. Worse, you can tell everyone else has already heard this by how they are barely paying attention, reading email.
You have a concern. You raised it in Slack, in the PRD, in the hallway on Tuesday. Each time you have tried a different phrasing: “maybe they just don’t understand what I’m saying,” you think. You raise it again now, fourth time’s a charm: “The auth service isn’t going to hold up under—”
“Totally hear that.” She’s already typing. “Let’s take it offline.” The VP of Product is here and, in the presence of true nobility, the help are not permitted to speak unless spoken to. You have taken a dozen things offline in recent memory, and none of them have ever been discussed.
Sarah barrels on. A table showing deliverables is flashed with a column for the owner. The list is exacting and labyrinthian, and it’s the first time you’re seeing it. “So, here’s what Eng is going to deliver by Q3.” What Eng—not your name, just the role, the livery, the uniform that will go on being worn long after whoever’s currently inside it has been quietly replaced—is going to deliver, the verb servants get. You deliver a tray. You deliver a package. You deliver a message you didn’t write.
Top of the hour, the meeting wraps. Everyone looks at you expectantly. You say “sounds good” because the personal cost of dissent is too high. You mean it, almost. You have become very good at meaning things almost.
Are you the tail wagging the dog or the dog wagging the tail?
If you’ve spent any time around FAANG, the story above probably feels like a bizarre role reversal. At top tech firms, engineering runs the show. The general manager of any given product line is on the engineering ladder and was promoted up from the SWE ranks. Budgetary control and binding decision-making authority rest entirely with SWE leaders. PMs in this world must beg, borrow, and steal any scrap of authority they can get their hands on. The influence they wield comes from relationships… or, in rare cases, pure rational authority, wherein the product leader is simply smarter than their engineering counterpart and can win most arguments.
The engineering-product partnership is not equal. Not even a little bit. I got my first taste of what it feels like to be on the other side of this when I switched into finance from big tech.
At the low- and mid-tier of the finance world, your business analysts and product managers rule the roost (the PM:Eng ratios at these companies are completely fucked, like 1:3). For the first time in my life, I was sitting in executive decision-making forums and felt vestigial. It was incredibly clear the room had only a passing interest in my opinion on any given matter—that my input would be taken under advisement but that I had no role beyond consultation. Once an analyst or PM made a decision, my 100% compliance and acquiescence were assumed. Formal power doesn’t change this. As a VP, product counterparts came seeking my approval but never my input. Any question or feedback I offered was treated as a puzzle to be solved and then set aside so the original plan could proceed unmodified. My role was reduced to a headcount vending machine and a name that could be wielded like a weapon to force my team’s compliance.
Even within my narrow lane of ownership (technology), if I made a decision on my own about purely engineering matters like architecture, PMs would take great umbrage with the fact I had not sought their consent and alignment. Complex and emotionally charged interactions would unfold where I would be expected to justify myself—the PM’s right to oversight and control over all matters affecting “their product” was assumed and absolute. Oh, and by the way, literally nothing was safe from PM control; we also had “platform PMs” who owned the internal engineering systems.
“Holy shit, this is horrible,” I thought to myself. And then it dawned on me, “Is this how my product counterparts felt back at Google?” Yep. 100%. This was a very poignant maybe-we’re-the-baddies moment.
Whether it’s product dominating engineering or engineering dominating product, I have never witnessed a truly equal partnership. I know they exist somewhere, but it’s the equivalent of finding a four-leaf clover…in the middle of a hurricane…wearing a blindfold.
Mutually Assured Destruction
XFN doesn’t stand for “cross-functional,” it stands for eXtremely Fucking Nonsensical.
Like, here is this person who I rely on for substantially all of my most critical job responsibilities through daily, verging-on-hourly interactions… and despite that, 90% of these relationships are the most corrosive, disorienting, and baggage-laden in your life.
The PM’s career progression is almost entirely decided by what the EM ships, how quickly, and how well the system actually works. Effectively, none of these things are in the PM’s control.
The EM’s career progression is almost entirely decided by what opportunities the PM identifies, whether those materialize as “business impact” once delivered, and whether the Eng Dir/VP actually give a shit. Effectively, none of these things are in the EM’s control.
Each party has a distinct management chain with pairwise product-engineering relationships laddered all the way up the org chart…each in various states of disrepair leading to conflicting directives and proxy wars.
Neither party can succeed without the other, but behind closed doors, both would tell you emphatically that the other side is completely useless and obstructive and we would all be better off without them.
AI has gifted us a new battleground to play out this enmity, with PMs chucking AI slop PRDs over the wall and EMs vibe coding creaky piles of shit that barely work. Attention to detail has not just exited the building; it has boarded a plane to another continent and is halfway across the Atlantic.
There’s no clear decision maker…on anything. That PM might have 10 years of IC SWE experience and a CS PhD and will challenge your shitty architecture decisions. That EM might have founded 2 startups with successful exits and will veto a tepid, limp-wristed excuse for a product strategy. Try to escalate, and the next 3 layers of management have the same problem. Your nearest shared node in the org chart is some SVP who is so disconnected from the details you might as well roll a d20 to make the decision.
I know this characterizes the relationship as exclusively hostile, but that is basically the default: any slight misalignment whatsoever will present to either party as obstructing one from doing their job. The whole thing can be summarized as two people holding each other at gunpoint screaming, “Move and I’ll blow your fuckin’ face off!”
Someone is reading this thinking, “No! My XFN relationships are great, and we work so well together! This guy is just a Bad Person.” I am, yes, but also I’m glad you’ve figured out how to cope; congratulations—but that’s all it is: cope. I too can figure out how to keep the wild dog from biting my face off, but wait a minute—why is there a wild dog in the middle of the cubicle farm biting faces off in the first place??
Peeling the Onion
Functional silos are entire reporting chains (managers reporting to directors reporting to VPs reporting to…yet more VPs lmao) of the same job function. An EM and a PM that co-own the same objectives don’t actually share a common ancestor until you reach the SVP or the CEO.
Each layer will have a different political reality, and it will directly shape your quality of life. For example, when I worked on Google Drive, our product director was literal furniture, and everyone (PMs included) knew the only thing that mattered was convincing the Eng director….but that didn’t extend to the VP level. The Product VP and Eng VP had an effective, loving, co-equal partnership, and you had to win over both. Go up one more layer and the Eng SVP was the only person who mattered. Add in sister orgs and the confusion multiplied. There were product leads in these other orgs so powerful they could upend our entire roadmap with a single email, while their Eng counterparts were basically sitting in the corner eating glue.
Getting anything done meant navigating multiple layers of people who mostly hated each other. It’s like peeling an onion. The more layers you peel, the worse it stinks, and now for some reason you’re crying.
You might think, “That’s fine, I can just talk to my partner, and I’m sure we can work things out,” but the reality of working for different managers will thwart you. As much as we try to be our own person, success at BigCo largely means figuring out what your boss really wants. You are directly incentivized in the form of career growth/power to prioritize what they think over the collective opinion of your peer group. See where this is going?
This is why cross-functional relationships are characterized by so much manipulation. The distinction between influence and manipulation is that the latter is done without the other person’s interests at heart. Your career advancement at the expense of your peer’s, collective truth-seeking, or what’s right for the customer is the very definition of “not in their interest.” We’re so good at lying to ourselves that this doesn’t even cross our minds; we equate what will get us rewarded with what’s best for everyone.
To keep any group headed in the same direction, you need alignment one level up (at a minimum). What this means in practice is arming your boss with a viewpoint, facts/data/analysis to back it up, and sending them off to do battle with their peer directors/VPs/whatever. If you’re precocious, you could try directly influencing these organizational Aunts and Uncles…but more often than not all I got was confused looks, “Why are you talking to me? Is your Dad home?” Hurray for hierarchy! Culture eats strategy for breakfast and all that tired shit.
Til Death Do You Part
But to be fully couched within the comfort of a friend is a mode existence with severe implications. To please you perfectly she must understand you perfectly. Thus you cannot defy her expectations or escape her reach. Her benevolence has circumscribed you, and your life’s achievements will not reach beyond the map she has drawn.
—Jonathan Blow’s Braid
How strange it is that this person is basically your work husband/wife—whose abilities and compatibility with you will fully determine the extent of your accomplishments—and yet you have no choice in the matter?
Influencing this is a quintessential part of cross-functional hell.
My myriad bosses over the years have never failed to bring up the topic, “How is <some random PM> doing?” or “What do you think of <some bozo>?” where the context clues and tone make the subtext incredibly clear: “I want to get this person replaced and need ammo.” In a previous job, every 1:1 was basically a dirt-gathering session, and it was an open secret that when a PM got fired, it was because the Eng director wanted them gone. The crazy thing? That was probably rightly so, as otherwise we would have been sabotaged by deadweight.
I was blissfully unaware of these backchannels as a line manager, but once I became a manager-of-managers, that changed. Literally every single 1:1 with my product counterpart began with the weekly Airing of Grievances where I would hear about insane shit my reports said or some random engineer who—according to the PM’s exceptional technical judgment 🙄—seems to be “moving too slow.” I would, of course, return the favor by putting on blast the latest coked-up PRDs and product reviews that were total shit shows. My role shifted from building products to playing recess monitor, and I had no choice in the matter: abdicating this responsibility meant leaving my org’s success to chance.
Any leader worth their salt has to master the skill of backseat managing their cross-functional partner’s hiring and firing decisions. I’m sorry to burst your bubble. This is happening quite literally nonstop at every single layer of the org chart. Someone, somewhere is gossiping about you and it is directly and irrevocably altering the trajectory of your career. You cannot opt to ignore this if you ever want to be anything more than a low-level IC.
You’re being hired into a marriage where the average case scenario is you will at some point get divorced, and the custody battle over the kids (aka who gets to stay on the team and who gets reorged to Timbuktu) will permanently traumatize you.
In an alternate universe, product and engineering leads would be hired as couples. A matched set that you already know works well together and has established strong working habits, compatible worldviews, and knows how to have tough conversations. But no, we live in a world where hiring managers try to play matchmaker.
Breaking the Nash Equilibrium
So if you “get what you get”, how do you make things work? How do you navigate the criss-crossing layers of cross-functional dysfunction with their conflicting priorities, mismatched incentives, and mutually assured destruction?
As much as I’ve characterized this as a futile prisoner’s dilemma, I’ve had legitimately wonderful cross-functional relationships… relationships where 1+1 actually makes 3 because we amplified each other’s strengths, covered our weaknesses, and had unbreakable trust, which meant disagreement was safe and productive (miss you, Ryan, Hari, Liz, Andrew, Luke…may we cross paths again someday).
Getting there involves some luck—not much you can do with a totally incompatible personality, ethics, or life priorities. Still, assuming none of that gets in the way, I’ve learned a few things over the years which directly contribute to building the ideal product-engineering partnership.
Do what you say and say what you mean.
This one is pretty simple. Actually follow through on your promises and never use doublespeak. If you say you’ll prepare a report by Tuesday, then get it done by Tuesday and make sure it’s high quality. The bar is so low it’s subterranean, and this alone will differentiate you from your peers. In roles like product that are dependent on others to do a substantial amount of the actual building, you can quickly become everyone’s favorite go-to collaborator. In the early days of the relationship, 100% consistent reliability will open the door to deeper trust.
Furthermore, never mislead, selectively omit facts, or cover up what you really think. You are trying to demonstrate with your deeds that you are totally above board and trustworthy. This also means challenging directly and having some uncomfortable conversations, but believe me when I tell you that beats being nice all the time, covering up what you really think, and then surprising your partner at the last minute. I will take a disagreeable person who displays trustworthy competence and integrity over a smiling imbecile that just wants everyone to “feel good.”
Day Long Jam Session
Take your XFN partner for a day-long 1:1 offsite. I usually attach these to the formation of a new team, a reorg, or the initiation of a multi-quarter initiative. The agenda is fairly flexible and contextual, but it’s always built around deep diving and aligning on key principles, decisions, and actions for whatever Big Thing you’re about to embark on together. To make things more exciting, I also plan a couple “over the top” exercises like card sorting or journey mapping…design-sprint-type stuff. These activities rarely happen, either because they are considered too time-intensive or because the team is so unfamiliar with them that they would never work at scale. Still, the intimate 1:1 environment is perfect for working around all of the above. Throw in a long lunch break somewhere nice and a couple field trips to top it off (I love taking people to my bougie coffee house where they can try single-origin pourover for the first time).
Something really magical happens when the two of you get away from the bustle and the prying eyes, and with enough time to let conversations expand to fill the time they need rather than artificial 30-minute blocks—when you can drop your polished professional persona, the need to maintain consistency with past proclamations, and the demand that every meeting resolve productively to a set of decisions and action items. Real vulnerability and intimacy can occur, and with them, myriad opportunities to show you can be trusted with fragile ideas, self-doubts, and all the messy stuff we sweep under the rug when doing battle in the Corporate Coliseum. Each of you will get to know how the other person thinks, what they value, their fears, and their aspirations.
No one I know does this. I always get excuses like “We’re too busy,” or “I’m not sure I’m comfortable inviting my product partner to something like that,” or “OMG what would we even talk about all day?” But this little trick has a 100% success rate at taking relationships to the next level. If I ever found myself in an environment where something like this was impossible, or my product partner wasn’t willing to play ball, I would seriously re-evaluate if that was the place for me and probably leave shortly thereafter—it’s that essential to establishing bidirectional trust and alignment.
Word Vomit
You ever have something important to say but aren’t quite sure how to phrase it? Maybe you’re worried about implying implications and hurting the other person’s feelings…so you stay quiet. Important disagreements and frictions go unresolved, and the psychic burden of all those loose ends and quiet resentments builds until one of you bursts.
To solve this, the two of you agree to create a container—a safe space to air your concerns—where the other person agrees to hold off on their immediate reaction, not read too much into the exact phrasing, and listen to understand and collaboratively sort out what you really mean. It’s a bidirectional acknowledgment that “I don’t know what I’m trying to say, but I need to say something; let’s figure it out together,” and for both parties to explicitly probe and test conclusions/implications rather than assuming the meaning you inferred was the intended one.
When one of these moments arises, you trigger the container with a key phrase. Known for my elegance and way with words, I always use “I need to word vomit.” If the other person acknowledges, you proceed. If they don’t accept the invitation, then you need to wait for a better time—maybe tempers have flared, and they aren’t in a productive headspace.
If this sounds very intimate, that’s because it is. You are applying a technique straight out of couples counseling and need an extreme level of comfort with your XFN counterparts. As a result, this technique may be out of reach for many of you. That’s especially so if you pair up with 5+ XFN partners—Dunbar’s Number guarantees you won’t be able to maintain enough of these deep relationships. Every interaction you have will be shallow and transactional. It’s why hyper matrixed environments are truly, deeply, and irrevocably structurally fucked.1
Disagree in Private, Support in Public
Never disagree with your XFN partner in group settings—save the debate for your 1:1.
That doesn’t mean you have to go out of your way to offer unequivocal support and endorsement—if you truly disagree (maybe they surprised you with a hot take mid-meeting), then stay quiet or talk around the issue. Anything but challenging them directly in front of the group.
A public disagreement can be disastrous. At an individual level, it triggers major status and relatedness threats that trigger fight-or-flight responses. The group may include their superiors or your subordinates, which puts both their professional standing and credibility at risk. Worse, your subordinates will pick up on the misalignment and create downstream chaos.
Look, I’ll be the first to say this is the path of cowards and anyone who can’t handle public disagreement is weak, their bloodline is weak, and they will not survive the winter. Unfortunately, there are, like…10? companies on Earth where the culture will support this sort of behavior. Unless you work at Bridgewater Associates or some shit, this is not a winning strategy.
Influence over Control
This one is from Unlocking Leadership Mindtraps. It’s gooey, HBR-style fan fiction, but god I wish I read this much, much sooner in my career.
Ask yourself not “what do I control” but rather “what can I influence?” Control is an illusion. Even if you had zero XFN partners and absolute authority within your domain, you are limited by the extent to which you can program the meat machinery that constitutes your team. More power changes nothing—new CEOs frequently express surprise at how little control they have. They climb all the way to the top to find out there are always more stakeholders (the board, markets, customers) and more constraints (resources, people, culture).
Humans have an instinct to seek control over their immediate environment for safety, which helps explain why the desire to control is so strong. However, attempts to directly manage complex, interconnected outcomes such as company culture fail because these systems emerge naturally rather than by following orders. Enforcing strict rules often results in unintended consequences and increased stress when things go awry.
Especially once you mix in a dozen XFN partners, you must shift from dictating specific outcomes to providing direction. Then ask yourself what conditions or support you can create to help that vision come to pass. You will have to deal with implicit limits that come with this mindset. Some things (and people) are outside your realistic influence. All you can do is nudge the system towards a slightly more optimal course one day at a time.
Ask “What Do You Need From Me?”
In The Effective Executive, Peter Drucker points out that in knowledge work, a specialist’s output is basically useless until someone else picks it up—your job isn’t done when you close the task, but once you figure out who touches it next and what they actually need to make it work.
Try asking your PM, “What do you need from me, in what form, and by when, to do your job well?” in your next 1:1. They will completely short-circuit because nobody thinks like this. XFN relationships usually flow the exact opposite direction (“what do you owe me? why is this late? why do you suck so bad?”)
This flows in both directions, too—you have to shift how you view your coworkers and start asking what they are genuinely great at and how you can use that contribution to maximum effect.
Most companies are terrible at this; org charts and rigid job ladders dictate what someone does day-to-day. That’s how a brilliant engineer can end up wasting half their week writing status updates just because that’s what “Tech Lead” has meant historically.
If you can do both things—ask what people genuinely need from you, and deliberately figure out how to leverage their contributions—your relationship will stop being adversarial by default. You will achieve what no amount of “alignment meetings” or process tweaks ever will: real honest-to-god collaboration.
Liberté, égalité, fraternité
Dealing with matrixed orgs and cross-functional partners is awful. Why do you think “builders” (a mythical chimera of Engineer+PM+Designer+PMM all rolled into one) or going “founder mode” (ignoring all boundaries and just doing everything yourself) have captured the zeitgeist so completely? This shit sucks so bad that even these Twitter Thought Leader Acid Trips sound more appealing.
In 2026, things are only getting worse as now there is a whole layer of anxiety about how AI will enable “Engineers to reclaim the PMs job” or “PMs to replace Engineers” (let’s not even talk about designers and researchers; people are 100% convinced those job families are donezo, which is the stupidest shit I’ve ever heard). Things were already hard enough, but now it truly is a zero-sum game—someone is leaving in a body bag.
I think we’ll all find out real quick that it’s actually very hard to be an excellent strategist, product designer, marketer, backend developer, and SRE all at once. Especially not when quality/craft increasingly matter for winning in the market. In the (admittedly rare) times in my career when our XFN cohort was able to move in unison towards a shared goal, the things we built were truly incredible—among the proudest accomplishments of my career.
The tips shared here—reliability, radical honesty, maintaining a shared front, creating containers for disagreement, letting go of control, and playing to mutual strengths—can help you recreate those winning environments, but there are limits. You can’t just “be super agreeable” and suddenly expect everything to work out. There will be cultures and people you can’t work with, and rather than make everyone suffer, you need to raise your hand and opt out. When interviewing for a new job, you must also demand to interview your XFN partners and be brutally honest with yourself: do you want to interact with these people every day for the rest of your career? This is even more important than assessing whether you vibe with your manager and your team.
Have fun out there and play nice everybody 😊
Find me a legacy company with diffuse responsibility operating in an environment ripe for technological disruption, and it’s just a matter of time until they stumble. I have made a killing shorting the stock of companies run this way; it’s literally like shooting fish in a barrel. They have no capacity whatsoever to respond to competitive threats, paradigm shifts, and market disruption.