How to Lead Through Outcomes Instead of Tasks
The advice to lead through outcomes rather than tasks is correct and almost always delivered without the mechanics, which makes it unusable. Founders who take it seriously typically replace specific instructions with vague ownership — “you own the launch,” “you own growth” — and get worse results than before, because a vague outcome is not more freedom, it is less information. This page is the mechanical version: what an outcome actually is, how to write one someone can be held to, and the three things you have to supply alongside it.
What’s inside
Outcome, output, activity
How to write an outcome that works
The three things you must supply
What to do when the outcome is missed
Where outcome-based leadership fails
The founder’s part
Where Blomma fits
Outcome, output, activity
Three levels, routinely conflated, and the confusion is the source of most of the trouble.
Activityis what someone does: run five experiments, make forty calls, ship the redesign.
Outputis what gets produced: a redesigned onboarding flow, a published report, a shipped feature.
Outcomeis what changes as a result: activation rate up, sales cycle shortened, churn reduced in a segment.
Most founders believe they manage outcomes and actually manage outputs. “Ship the new onboarding by March” is an output with a date, and it can be fully achieved while the thing you wanted — more people getting to value — does not happen. The person has done exactly what you asked and you are unsatisfied, which is a bad position for both of you and entirely your doing.
The practical difference: an output tells someone what to build, so they build it. An outcome tells them what to achieve, so they can tell you the thing you asked for will not work and propose something better. That second conversation is the entire return on outcome-based leadership, and you only get it if the outcome, not the output, is what you named.
How to write an outcome that works
Four criteria. An outcome that misses any of them will not hold under pressure.
It names a change, not a deliverable.“Activation rate from 22% to 35%” rather than “new onboarding shipped.” If a person could satisfy it without anything improving, it is an output.
It is measurable, and you both agree how.Not necessarily a number — some outcomes are qualitative — but the measure has to be agreed in advance. The most common failure in outcome-based leadership is discovering at the review that you were each tracking something different.
It is genuinely within their influence.This is the criterion founders violate most. Giving someone an outcome whose main drivers sit with three other teams sets them up to fail and to spend their time on negotiation rather than work. Check the drivers before you assign it.
It has a horizon.A quarter, a half, a year. Outcomes without time bounds cannot be reviewed, which means they cannot be owned.
Then apply the test that catches most bad ones:could two reasonable people disagree about whether this was achieved?If yes, it is not written well enough, and you will have that disagreement at the worst possible time.
The three things you must supply
An outcome without these three is not delegation, it is abandonment dressed as trust — and it is why so many attempts at this produce worse results than task management did.
The decisions.If they own an outcome, they need the decisions that drive it. Someone owning activation who cannot change the onboarding flow, the pricing, or the sequencing owns a number rather than an outcome. Trace the drivers and hand over the corresponding calls, or narrow the outcome to match the authority.
The context.Why this outcome matters now, what has been tried, what the real constraints are, what you know that they do not. Founders assume context is obvious because they have been carrying it for years. It is not, and its absence is why capable people pursue outcomes in ways that surprise you.
The standard.What an acceptable path looks like. Outcome-based leadership does not mean the method is entirely open — there are usually things you would not accept, like hitting activation by degrading quality or hitting revenue by discounting the base. Say those out loud in advance. Discovering your unstated constraints after the fact is the specific experience that makes people conclude the autonomy was fake.
What to do when the outcome is missed
This is where outcome-based leadership either establishes itself or collapses back into task management.
Start with the reasoning, not the result. “Walk me through how you approached it” tells you whether this was a judgment problem, an execution problem, an information problem, or an outcome that was wrong to begin with. Those are four different situations with four different responses, and you cannot tell which from the number.
Separate a missed outcome from a bad process. Someone can run an excellent process and miss, because outcomes have variance and some bets do not land. Someone can hit an outcome through luck or by doing something you would not have sanctioned. Reward the process, review the result — if you only reward the number, you will get number-hitting behaviour that you will not like.
Check whether the outcome was well-set before you assess the person. A significant share of missed outcomes were unachievable, or depended on drivers the owner did not control, or were measured differently by each party. That is your error, and treating it as theirs is how you lose good people.
Then decide what changes, specifically, and write it down. A missed outcome that produces no change in approach will be missed again.
Where outcome-based leadership fails
Three situations where it is the wrong tool, worth knowing so you do not apply it universally.
With someone genuinely new to the work.A person who has never done this needs more direction, not less. Handing a novice an outcome and standing back is not respect, it is a setup. Move to outcomes as capability builds.
Where the outcome is genuinely outside anyone’s control.Some work is real and does not map to a moveable metric — early research, platform investment, some compliance work. Forcing an outcome frame produces invented metrics, and invented metrics get gamed.
In genuine crises.When something is on fire, direct instruction is appropriate and everyone understands why. The failure is not returning to outcomes afterwards, so the crisis mode becomes permanent.
And one broader caution: outcome-based leadership requires you to tolerate methods you would not have chosen. If you cannot, you will keep intervening on the how, and your team will correctly conclude that the outcome framing was decoration.
The founder’s part
Two behaviours determine whether this works, and both are yours.
The first is not intervening on method when the outcome is on track. If someone is hitting the number in a way you find inelegant, that is the cost of the model and it is cheaper than the alternative. Every intervention on method teaches people that the real standard is your preference, and they will optimise for it.
The second is holding to the review rather than the running commentary. Outcomes are reviewed at a cadence. Founders who set an outcome and then ask about it weekly have created a task-management relationship with an outcome label on it, and the person on the other end can feel the difference precisely.
One test of whether you have made the shift: in your last month of one-to-ones, did you spend more time on what people were doing or on what was changing? Most founders who believe they lead through outcomes discover the answer is the former.
Where Blomma fits
The difficulty here is not the framework, it is the writing and the restraint. Writing an outcome that is measurable, within someone’s influence, and not secretly an output is harder than it looks, and most founders’ first attempts fail the four criteria. Then the harder part: watching someone pursue an outcome by a route you would not have taken, and not saying anything.
Blomma is an always-on AI career coach with no stake in your company. Use it to draft the outcomes for your executives or team leads and test them against the four criteria — particularly whether the drivers actually sit with the person. Use it to surface your unstated constraints before they cause a problem, which is the most useful ten minutes in this whole exercise. Use it to prepare the missed-outcome conversation so it starts with reasoning rather than the number. And use it in the moment you want to intervene on method, to work out whether the objection is about the outcome or about your preference.
The second application: the people you are handing outcomes to are being asked to exercise judgment they have not needed before. Coaching is what closes that gap fast enough for the model to work..
Leading through outcomes is not leading vaguely. It is being far more precise about what you want to change and far less prescriptive about how — and supplying the decisions, context, and standard that make the second part possible.