Hold is the most misread competency in Resolve. People hear “holding a position” and think it's about stubbornness — the PM who digs in, won't listen, wears conviction as armor. That's not what we're measuring. We're measuring something narrower and harder: the ability to stay with a well-reasoned call under social pressure without either caving to get the meeting to end or overplaying your hand to appear more certain than you are.
The two failure modes are symmetric. Fold too easily and your judgment ceases to matter — you become a vote-counter, not a decision-maker. Bluff with false certainty and you erode trust the moment reality contradicts you; worse, you model a culture where confidence is performed rather than earned.
What hold specifically tests — and what no other MARK competency does — is the gap between when you stop updating and when you start defending. Signal measures whether you read evidence correctly. Power measures whether you'll say an uncomfortable truth to someone with authority over you. Hold measures what happens after those calls: can you stay with the position you reached through good judgment, when the social cost of staying rises, without either abandoning it to reduce friction or fabricating certainty you don't have?
The wrong beginner take: “I just need to be more confident.” Confidence is a surface property. Hold is about the integrity of the gap between what you believe and what you say — in a room full of people who want you to say something different.
citation
PL Standard v3.1 · Resolve · Hold
Markdown form: [PL Standard v3.1 · Resolve · Hold](https://pragmaticleaders.io/framework/competencies/hold)
the four levels
Anchored at every rung. The blockquote at each level is panel-authored and pulled live from the rubric — edit anchors via the panel tooling and they appear here.
L1Developing
folds under the first credible challenge — rephrases the position as a hypothesis or defers to the objector.
you'll see this when…
The PM revises their recommendation mid-meeting when someone senior expresses doubt, without new evidence entering the conversation.
They hedge retroactively — presenting a call in the brief, then walking it back in the review to “let the team decide.”
They conflate social comfort with updated belief: “I changed my mind” turns out to mean “I didn't want the conflict.”
common failure mode: The PM treats consensus as validation — if the room isn't agreeing, the call was probably wrong.
L2Competent
holds a stated position under routine pushback; escalates to a manager before standing firm when stakes rise or a senior stakeholder objects.
you'll see this when…
The PM holds the position through the first round of pushback but starts softening the edges after the second.
They can name why they held (“the data still points here”) but can't always separate that from social exhaustion — sometimes they hold because the meeting ended before the pressure peaked.
They introduce unnecessary hedges that lower the practical value of the call: “obviously there's uncertainty here” said in a way that signals retreat, not calibration.
common failure mode: Holding on the record but conceding in the hallway — the position survives the meeting, then quietly dissolves in the follow-up Slack.
L3Proficient
holds a hard call under sustained pushback from multiple stakeholders, names what evidence would change the view, and stays on the position in the room — not in the follow-up note.
you'll see this when…
The PM can name, in the room, exactly what would change their position — and it isn't “if someone disagrees loudly enough.”
They distinguish between updating on new evidence (correct) and caving to pressure (not), and they make that distinction visible: “that's not new information, it's a preference.”
They hold without escalating — the room stays functional; they're not defensive, not distant, not martyred.
common failure mode: The PM holds the call but loses the room — correct in substance, corrosive in delivery; the position survives but the relationship doesn't.
L4Expert
holds the call when everyone else has folded, names the failure mode that will happen if the team pivots, and absorbs the social cost of being right publicly.
you'll see this when…
The PM can hold a position that has already been overruled — they document the call, note their dissent, and execute the decision without undermining it, keeping the record clean for when reality lands.
They know when not to hold: they've developed the meta-judgment to distinguish genuine new evidence arriving late from the same old pressure in a new disguise.
They create conditions where others feel safe holding their own positions — by modelling that updating on evidence is valued and caving to rank is not, they change what's possible in the room.
common failure mode: The expert holds so consistently and visibly that they become known for holding, which starts to function as its own social pressure — junior PMs stop bringing contrary evidence because they assume the position is already fixed. The L4 trap is turning a virtue into a chilling effect.
how to develop it
The most leveraged move is building a sharp personal heuristic for why your position is changing when it changes. Most PMs don't track this. The question is simple: is new information entering, or is social pressure entering? Keeping that distinction live — in the moment, not in retrospect — is the practice.
Read. Start with /manual/presenting-to-leadership — not because leadership presence is the goal, but because it breaks down the specific dynamics where hold gets tested most visibly. Then /manual/data-informed-decisions for the structural counterpart: what it actually looks like to update on evidence versus update on discomfort. /manual/roadmap-prioritization has a section on holding prioritization calls under feature-request pressure worth reading alongside this.
Practice. The scenarios that stress-test hold most directly are the ones where pressure is social, not technical. Take a Practice session where you've made a call with data and a credible stakeholder says no. Run it until “I need to think about this” stops feeling like the right answer.
Write. Draft a Brief response where your recommendation is explicitly against what the team wanted to ship — something you'd have to defend in a product review. Don't soften the edges in the write-up to pre-empt the pushback. Write the call as you actually believe it. Then notice where you hedged when you didn't need to.
Coach yourself. After your next product review, ask: did my position at the end match my position at the start? if not, what changed — and was it evidence or atmosphere? That's the journal entry. One sentence. Keep it weekly for a month.
how to spot it in others
In 1:1s: when they recap a recent call that went against them, do they say “the room decided X” or “I ended up agreeing” — and can they tell you which one it actually was?
In planning: when a feature request comes in above their prioritization line, do they re-rank to accommodate, or do they explain the existing logic and hold the line pending a real tradeoff?
In post-mortems: when a call they made turned out wrong, do they distinguish between “I held it too long past the evidence” and “I was right to hold but the outcome was bad” — or do they collapse those two?
In all-hands: do they present ambiguous situations as more certain than they are? The tell isn't swagger — it's the absence of the calibration language that would make the uncertainty visible.
After being overruled: do they execute cleanly, or do they passively let the decision fail — death by malicious compliance?
three failure modes we see often
The consensus fold. The PM reads agreement as validation, so they chase it. Every meeting that ends without full buy-in feels like a signal to keep revising the position. The cost is subtle at first — it looks like responsiveness, like being a good listener. The tell is that the position never stabilizes; every new conversation is an opportunity to re-discover what the room wants. Over time, no one trusts the PM's read, because the read keeps moving.
The credentialing bluff. The PM doesn't have enough genuine conviction to hold under fire, so they perform it instead — asserting certainty they don't feel, citing data with more confidence than the data supports, invoking authority (“I've seen this fail at two other companies”) to shut down debate. It works until reality contradicts them, at which point the trust damage is disproportionate: they weren't just wrong, they seemed certain. The bluff makes every future position suspect.
The martyrdom hold. The PM holds their position, but makes holding it a dramatic act — sighing, noting their dissent loudly, reminding the room that they called this in the debrief three months later. The position was right, but the way they held it turned conviction into a personality trait rather than a practice. The cost is relational: people stop engaging with their calls on the merits because every disagreement feels like it's going to become a story they'll have to hear again.
Power is about telling someone with authority over you that they're wrong. Hold is about what happens after that moment: sustaining the position across repeated pressure over time, including from peers and your own discomfort. Power is the first cut; Hold is the stamina. You can be strong on Power (willing to say the hard thing once) and weak on Hold (walking it back across three follow-up meetings). They're often tested together but they're distinct skills.
Miss is what you do when a call you held was wrong and the evidence is in. Hold is about the period before the evidence is in — sustaining a judgment through social pressure when you don't yet know the outcome. The failure modes look opposite: someone weak on Hold folds before they should; someone weak on Miss doesn't fold after they should. MARK measures both separately because being good at Hold doesn't protect you from refusing to own a clean miss.
Signal measures whether you read conflicting evidence to the right conclusion. Hold assumes you read it correctly and tests what you do with that read when it becomes socially costly. Someone can be a strong Signal reader and a weak Hold — they know what the data says, but they cave when the room disagrees. Improving Signal without Hold produces PMs who see the right answer but can't sustain it.
what good looks like in the wild
A product team is six weeks from launch on a feature they've built for eight months. The PM called a cut in the scope four months earlier — a specific integration that two of the company's largest accounts had been asking for. The call was documented: the data showed adoption wouldn't move even with the integration in place, and adding it would slip the date. The CPO agreed at the time. The design lead agreed.
Three weeks out, one of the accounts hears through the sales team that the integration isn't coming. They escalate. The account executive comes to the PM in the week before a product review with a detailed argument: this account represents 18% of ARR, they've been asking for this since Q3, a competitor just shipped something adjacent, and the relationship is at risk. The PM listens to all of it. She asks the AE to send her the exact conversation that triggered the concern.
In the product review, the CPO — who had agreed to the cut four months earlier — opens by saying the scope call may need revisiting. The PM stands up what she found: the account conversation was real, but the competitor feature wasn't adjacent — it was solving a different problem. The 18% ARR figure was current ARR; the integration would have affected renewal probability at best, not expansion, and the account's NPS had been declining before the feature ask. None of this was new data; it was the same data, now arriving under pressure.
She held. She also did one thing that separated the hold from stubbornness: she named what would change her mind. “If the account gives us written confirmation that this is a renewal blocker — not a preference, a blocker — I'll reopen the scope call today.” No one could produce that. The feature shipped without the integration. The account renewed. The CPO noted in the next planning cycle that the call had been right, and that the way it was held was what let the team trust the call rather than just comply with it.
“
That's what Hold looks like when it's working — not the absence of pressure, but the ability to stay with evidence while the pressure rises, and to make the criteria for changing visible so the hold is legible, not just stubborn.