🎯 Goals, Constraints, and Success Criteria

By Better Wise

In the previous lesson you compressed a messy situation into one sentence: the real decision, not the event that triggered it. That sentence still leaves the most important question open. What would a good answer have to do? This lesson sets that bar, so you stop arguing about options before you agree on what counts as a win.

By the end of this lesson you can:

  • turn a vague wish into a goal with a direction, a horizon, and a named beneficiary
  • sort limits into hard constraints, soft costs, and disguised preferences
  • write success criteria that two people who disagree could still grade the same way
  • draft the goals and constraints block of your decision brief

From a wish to a ranked goal

A wish gives you nothing to trade against. "I want more interesting work" cannot be failed, compared, or reviewed. A goal adds three things: a direction (more autonomy, less travel), a horizon (by next June), and a beneficiary (me, my team, my family). Add those and the option set starts to shrink on its own.

Example: Priya, an operations manager, starts with "I want more interesting work." Reframed: "Within 12 months, own at least two projects end to end, with no drop in income." That version can be compared against a job offer, an internal transfer, and staying put.

Ranking is where most frames quietly fail. If autonomy is first and income stability second, a lateral move with a small pay cut is a live option. If you refuse to rank, every candidate that scores well on anything looks acceptable, and the real trade-off stays hidden until you are already committed.

Sign in to view the full content of this lesson. Sign in

Answer guidance for Exercise 1

Exercise 1. Rewrite each wish as a goal, then rank the two against each other.

  1. "I should get better at public speaking."
  2. "Our team needs to stop missing deadlines."

A workable pair: (1) "Deliver three internal talks in the next six months without reading from notes." (2) "Ship 90% of sprint commitments by Q3, measured weekly." Ranking depends on your situation: if the deadline failures are costing client trust, that goal outranks your speaking practice. There is no universally correct order. There is only an order you can defend.

Constraints: hard lines, soft costs, and disguised preferences

A constraint is not a goal. A goal pulls you forward; a constraint fences the field. The distinction that does real work is hard versus soft.

Type Behaviour in the comparison Test question
Hard constraint Violation removes the option entirely "Would I reject this even if it were excellent on everything else?"
Soft constraint Costs points, can still be traded "What would I give up to satisfy this?"
Preference Shapes satisfaction, not eligibility "Is there a real limit behind this, or a taste?"

Hard constraints need a source and an owner: a legal requirement, a budget ceiling, caregiving hours, a medical limit, a contractual notice period. If you cannot name the source, you probably have a soft constraint.

Two failures show up constantly. Constraint inflation declares almost everything non-negotiable, which empties the option set and makes the decision look impossible. Constraint laundering dresses a preference as a rule so the conversation ends without evidence. "I will not consider relocating" is a hard constraint if a lease or a parent's care makes it one, and a preference if it is really about discomfort.

A condition that removes an option no matter how well it scores elsewhere is a _____. A limit that costs an option points but can still be traded is a _____, and dressing a preference up as a rule to end the debate is _____.

Answer guidance for Exercise 2

Exercise 2. Sort into hard constraint, soft constraint, or preference:

  1. "The role must be within 45 minutes of my mother's care home."
  2. "I would rather not work Fridays."
  3. "Total compensation cannot fall below $95,000 because of my mortgage."
  4. "I want a team that is not chronically firefighting."

Likely answers: 1 hard (caregiving limit with a physical source). 2 soft or preference, depending on whether Friday hours are a real limit or a comfort. 3 hard while the mortgage runs, and worth re-testing at renewal. 4 soft, and worth converting into a testable criterion later. Your own sorting is the one you have to defend.

Success criteria that survive hindsight

A success criterion is a statement you can check after the fact. Four properties make that possible: it is observable (someone else can verify it from evidence), time-bound (a date), threshold-based (a number or a yes/no event), and tied to a named goal.

Separate process criteria (what you will do: run four interviews, prototype with three users) from outcome criteria (what should be true later: two projects owned end to end). Process criteria keep you honest while the decision is running. Outcome criteria decide whether it worked.

The stranger test: could two people who disagree about the decision grade the result the same way from the same evidence? "I felt more engaged" fails. "My manager confirms in writing that I own two projects end to end by March" passes.

Write the criteria before you choose. Pre-registering them is the cheapest protection you have against moving the goalposts and against grading a decision purely by how it turned out.

A testable success criterion must be observable, time-bound, threshold-based, and _____. Writing criteria before you decide protects you from _____ when results arrive.

Answer guidance for Exercise 3

Exercise 3. Take one criterion you would normally write, and fix it.

Weak: "The new process should make the team happier." Fixed: "By 30 September, at least 8 of 10 team members rate workload as manageable or better on the quarterly survey, up from 4 of 10." Notice what changed: an instrument, a date, a threshold, and a named goal. Strengthen any criterion by adding those four elements, not by adding adjectives.

Assemble the frame: your one-page bar

flowchart TD
  W["🌫️<br/>Wish"] --> G["🎯<br/>Ranked goals"]
  G --> HC["🚧<br/>Hard constraints"]
  HC --> O["⚖️<br/>Options compared"]
  O --> SC["📏<br/>Success criteria"]
  SC --> R["🔍<br/>Review date"]

Keep the frame to a page. It travels with the decision.

Decision (one sentence):
Ranked goals:        1) ______  2) ______  3) ______
Hard constraints:    ______ (source: ______)
Soft constraints:    ______ (tradeable for: ______)
Success criteria:    primary (observable, date, threshold)
                     secondary
                     kill criterion: ______
Open assumptions:    ______ (to test in the next lesson)

Answer guidance for Exercise 4

Exercise 4. Fill the template for your project decision. Then run two checks: (a) delete the top-ranked goal and see whether your shortlist changes. If it does not, your ranking is decorative and needs another pass. (b) Hand the success criteria to someone who disagrees with your plan and ask them to grade a hypothetical outcome. If they cannot, rewrite the criteria.

Traps that quietly reset your bar

  • Ranking by importance instead of by trade-off. Ask which goal you would sacrifice first when two conflict. That answer is your real ranking.
  • Constraints with no source. If you cannot name who or what enforces it, move it to soft constraints and watch what happens.
  • Criteria that only measure effort. "I worked hard on this" is not a result. Pair every process criterion with an outcome criterion.
  • Criteria written after the result. If you find yourself defining success in the past tense, you are explaining, not evaluating.
  • Too many goals. Three ranked goals is a usable bar. Seven is a wish list wearing a suit.

What to carry forward

You now have a bar: ranked goals that force trade-offs, constraints sorted by whether they truly eliminate an option, and criteria a stranger could check. The frame is only as good as its inputs, though, and right now your inputs are a mix of facts, preferences, and assumptions that all feel equally solid. The next lesson, Facts vs Preferences: What Counts as Evidence, sorts them.

Front: What turns a wish into a goal? Back: A direction, a time horizon, and a named beneficiary, so trade-offs can be judged.

Front: Why must goals be ranked? Back: Unranked goals make every option that scores on anything look acceptable and hide the real trade-off.

Front: Hard constraint Back: A condition where violation removes the option entirely, however well it scores elsewhere.

Front: Soft constraint Back: A condition that costs an option points in the comparison but can still be traded away.

Front: How do you test whether a constraint is truly hard? Back: Ask whether you would reject an otherwise excellent option for violating it.

Front: Constraint inflation Back: Declaring almost everything non-negotiable, which empties the option set and hides the trade-off.

Front: Constraint laundering Back: Dressing up a preference as a rule so the debate ends without evidence.

Front: What must a hard constraint have? Back: A named source and an owner, such as a law, a budget ceiling, or caregiving hours.

Front: Four properties of a testable success criterion Back: Observable, time-bound, threshold-based, and tied to a named goal.

Front: Process criterion vs outcome criterion Back: Process describes what you will do; outcome describes what should be true afterwards.

Front: The stranger test for criteria Back: Two people who disagree should be able to grade the outcome the same way from the same evidence.

Front: Why pre-register success criteria? Back: To stop goalposts moving when results arrive and to reduce grading a decision by its outcome alone.

Front: Kill criterion Back: A pre-agreed observable signal that ends or reverses the decision, set before you start.

Front: What belongs in the goals and constraints block of a decision brief? Back: One-sentence decision, ranked goals, hard and soft constraints with sources, testable criteria with a date, and open assumptions.

Front: Which is a preference: a $95,000 pay floor or a dislike of Friday work? Back: The Friday dislike, unless a real limit such as contract hours or caregiving makes it hard.