Navigating Competing Priorities with Peers When Nobody Outranks Anybody

The platform team froze deployments for two weeks to run a database migration. That same week, the product team promised its largest customer a fix inside ten days. Both leads had receipts – the migration had already slipped twice, the customer was already angry. Neither could overrule the other, and the VP they shared was two levels up and booked until Thursday.

Competing priorities between teams get resolved one of three ways: somebody escalates and a boss picks a winner, somebody quietly caves, or the two leads negotiate the trade-off themselves against criteria they both accept. The first is expensive, the second is invisible until it isn't, and the third is a learnable skill with a specific sequence – acknowledge the other side's constraint before arguing your own, make the trade-off explicit instead of relitigating whose goal matters more, and agree on decision criteria before either of you defends a position. Escalation stays available, but as a move you make together, not a card you play against each other.

That sequence is the whole article. Here's how each piece works.

Why is peer conflict the hardest kind to resolve?

Because nobody outranks anybody. Disagreeing with your boss is uncomfortable, but the structure is clear: you make your case, they decide, everyone knows the rules (I've written about how to do that well). Disagreements inside your own team resolve the same way, with you in the deciding chair. Peer conflict between teams has no chair. Two people with equal authority, accountable to different metrics, holding commitments neither of them invented, have to build consensus under conditions where neither has complete information.

Most team conflict examples in training catalogs stage a manager and a direct report, which is a shame, because the peer version is more common and harder. In the map of conflict types product teams actually face, dependency and priority fights sit squarely in this territory: "We're blocked on the platform team. Again."

The instinct, when there's no tiebreaker, is to go find one. Which brings us to escalation.

What does escalating to a shared boss actually cost?

More than the meeting invite suggests. Three costs, in ascending order of damage.

The first is decision quality. Your boss has less context than either of you, and whatever they decide will be faster than it is good.

The second is the relationship. Unilateral escalation tells your peer you'd rather beat them than solve it with them. You'll still be working with this person next quarter, and the next fifty dependencies between your teams will be negotiated in the shadow of this one.

The third cost shows up after the decision, and it has research behind it. Hart and Schweitzer studied what happens post-agreement in work with a title that says it all: "Getting Less: When Negotiating Harms Post-Agreement Performance." You can win a negotiation – force the other party to accept the deal – and still receive less value than you expected, because a counterparty who feels bullied into a bad outcome underperforms during implementation. A peer whose priority got knocked down through a process that felt coercive may comply with the decision while quietly withdrawing discretionary effort from everything adjacent to it. You won the argument and lost the execution.

Start by acknowledging their constraint, out loud

The counterintuitive first move isn't stating your case. It's stating theirs.

Before arguing for your priority, describe their constraint accurately enough that they'd sign off on your summary: "You've slipped this migration twice, and the risk gets worse every week it waits. The freeze is you protecting the whole platform." Then check – did I get that right?

Two findings from the research back this move. First, listening isn't agreement. You can state their constraint completely without conceding anything, and the act of doing so drops defensiveness enough that they become capable of hearing yours. Second, researchers distinguish perspective-taking – imagining what the other person thinks – from perspective-getting, which is just asking them. Asking wins, and by a wide margin, because of something called the privileged knowledge problem: you can't set aside what you know when you imagine someone else's view, so you systematically overestimate how much their picture matches yours. When two leads sit under different metrics, different dashboards, and different executive pressure, inference fails quietly. Questions don't.

My co-designer on Conflict Campaign, Adam White, is a licensed professional counselor who spent 13 years as an electrical engineer before changing careers. His favorite phrase in a stuck conversation is "Oh, that sounds like new information to me." It's disarming because it's honest – it announces that your model of the situation just changed, and it invites the other person to keep going. (More phrases that open conversations instead of closing them.)

Make the trade-off explicit instead of relitigating the goals

Watch two team leads argue for an hour and you'll notice most of it is spent relitigating goals neither of them set. Should marketing have committed to that date? Should the migration have been scheduled this quarter? Those questions have no live options attached. The goals are set, the constraint is real, and the only actual decision is the trade-off.

Take the classic version: marketing has bought ad placements around a launch date, and engineering's honest readiness estimate runs two weeks past it. The unproductive conversation is about whose commitment is more legitimate. The productive one is a single sentence: we can ship on the date with two known defects and a patch plan, or slip two weeks and eat the campaign spend – which risk do we want? Everything before that sentence is conflict. Everything after it is decision-making.

Explicitness matters more than it feels like it should because of loss aversion. Losses loom two to three times larger than equivalent gains, which means your concession always feels bigger to you than their concession does – even when the two are objectively balanced. Left implicit, both sides walk away feeling like the loser. Named and written down, the trade-off shrinks to its actual size. And when the cost really is lopsided, say so: "I know this slip is harder for you than for us." Research on structuring agreements finds that acknowledging asymmetry builds acceptance rather than weakening your position.

When Adam and I wrote the design objectives for Operation Aetherfall, our pilot training scenario, we compressed this entire skill into three observable behaviors: builds consensus under uncertainty, explicitly acknowledges trade-offs and unknowns, gathers input from everyone at the table. Across two full pilot runs, the moment a stuck table came unstuck was almost always the moment someone said an unknown out loud instead of arguing around it.

How do you negotiate when both priorities are legitimate?

Against shared criteria, agreed before either of you defends an option. This is Fisher and Ury's principled negotiation applied to the cross-team case: when interests genuinely collide, decide on standards rather than stamina. (The upstream skill – separating positions from the interests underneath them – is covered in interest-based problem framing.)

The sequence:

1. State each other's constraints until the other side confirms your summary is accurate.
2. Name the trade-off in one sentence you both accept: "We can have X or Y by this date, not both."
3. Agree on decision criteria before evaluating options – customer impact, revenue risk, cost of delay, reversibility. Criteria chosen after positions harden become weapons.
4. Score the options together, flagging unknowns explicitly instead of papering over them.
5. Write the decision down – what's deprioritized, until when, and what new information would trigger a reassessment.

Steps 3 and 5 carry the most weight, and procedural justice research explains why: people accept outcomes that disadvantage them when the process feels fair – voice, consistency, transparency, neutrality, and the ability to revisit the call if facts change – more reliably than they accept favorable outcomes reached unfairly. The lead whose project waits until Q4 will implement the decision wholeheartedly if they helped choose the criteria that decided it.

There's a paradox buried in here worth naming. A computational study of team negotiation found that the side demonstrating willingness to negotiate ends up more influential over the final decision, not less – flexibility reads as strength. Refusing to come to the table doesn't protect your priority. It just guarantees the decision gets made somewhere with less context.

When should two peers actually escalate?

When the criteria conversation fails honestly – you agree on the facts and the options and still disagree about which risk the company should accept – or when the conflict has hardened past what a conversation can reach. Friedrich Glasl's nine-stage escalation model gives usable markers: in stages 1 through 3, the parties can still resolve things themselves; by stage 4, when the goal has shifted from solving the problem to limiting the other side's wins, outside help is required. Workplace mediation resolves disputes at an 80-85% rate, and it works best early, before positions calcify.

The mechanics of escalating matter as much as the timing. Done unilaterally, escalation is an ambush – your peer learns about it when their boss forwards the thread. Done jointly, it's two leads doing their jobs. Go together. Bring the shared framing, the agreed criteria, and the one genuine disagreement. Then ask your boss to make the weighting call, not to pick a winner: "We agree on the options and the costs. We disagree on whether the customer commitment or the platform risk should win this quarter. That judgment sits above both of us."

That framing converts the escalation from a verdict on two people into a decision about the work. The research on trust in peer conflict supports the joint version, too: even when trust between two parties is fragile, information from a third party about why a decision is fair is what enables collaborative follow-through.

Nobody is coming to referee the day-to-day between your team and theirs. The next fifty conflicts will go the way this one goes.