// the frameworks shelf

SMART goals, and the one thing they cannot catch.

The five letters as Doran actually wrote them, six examples rewritten, the honest critique, and where SMART sits alongside OKRs. Short version: keep it, and know what it is not looking for.

Written by Matt Roberts, ZOKRI co-founder Last reviewed August 2026 About an eight minute read
// the short answer

SMART is a test applied to a single goal sentence: Specific, Measurable, Assignable, Realistic, Time-related, in George Doran’s original 1981 wording. It is excellent at the sentence, and silent on everything above it.

SMART writes a good sentence. A goal system decides which sentences are worth writing.

It is not a rival to OKRs. It is a component already running inside them: every well-written Key Result is a SMART sentence. Teams fluent in SMART arrive at goal writing with the hygiene pre-installed, which is a real head start.

// the five letters

What each letter is actually asking, and how it gets read instead.

Doran’s original wording is in the left column. Most of what goes wrong with SMART in practice happens in the right one.

Doran, 1981 What it is really asking Where teams get it wrong
S
Specific
“Target a specific area for improvement.”
Which area, named precisely enough that two people would point at the same thing. Specific about the activity rather than the change. “Run a campaign” is perfectly specific and completely empty.
M
Measurable
“Quantify, or at least suggest, an indicator of progress.”
Can we see movement in time to act on it, from a baseline we actually hold. A measure chosen because it is easy to collect rather than because a decision depends on it.
A
Assignable
“Specify who will do it.”
One name, accountable, not a department. Quietly replaced by “Achievable” in later variants, which changed the framework: Assignable was about ownership, Achievable is about ambition, and it argues the other way.
R
Realistic
“State what results can realistically be achieved, given available resources.”
Is this target credible with what we actually have. Read as permission to lower the number rather than as an instruction to resource the goal properly.
T
Time-related
“Specify when the result can be achieved.”
By when, and how we will know we are late before we are late. A date taken from the calendar rather than from the work, so every goal becomes quarter-shaped whatever its real length.

Later variants read Achievable and Relevant in place of Assignable and Realistic. Doran wrote Assignable, and the swap is not cosmetic: one criterion is about ownership, the other about how much ambition is permitted.

// six examples

Five rewritten, and one that passes anyway.

The last one is the point of the whole page. Read it twice.

Not a goal yet

Improve customer onboarding.

SMART

Raise the share of new workspaces activated within 14 days from 38% to 60% by 30 September, owned by Sam.

Not a goal yet

Get better at sales.

SMART

Raise win rate on qualified opportunities from 24% to 33% by the end of Q3, owned by the sales lead.

Not a goal yet

Reduce the support load.

SMART

Cut support tickets per new customer in month one from 4.1 to 1.5 by 31 October, owned by the support lead.

Not a goal yet

Be more visible in the market.

SMART

Raise the share of new pipeline sourced from owned channels from 18% to 35% by 31 December, owned by the marketing lead.

Not a goal yet

Become a better manager.

SMART

Raise my team’s “my manager gives me feedback I can use” score from 54% to 75% by the November survey.

Not a goal yet

Launch the mobile app.

Flawlessly SMART, and still wrong

Launch the mobile app by 30 September, owned by the product lead.

Specific, assigned, dated, realistic. Also a task: it can ship in full while nothing changes. SMART cannot see this.

The pattern worth memorising

[Measure] from [baseline] to [target] by [date], owned by [name].

Then one sentence more, which SMART does not ask for and which changes everything: why that measure, and why that number. Without it, next quarter nobody can tell a stretch from a guess.

// the honest critique

Six limits, in the order they cost you.

01

It cannot tell a task from an outcome

“Launch the mobile app by 30 September, owned by the product lead” is flawlessly SMART. It is specific, measurable in the trivial sense, assigned, realistic and dated. It is also a task: it can be delivered in full while nothing about the business changes. SMART will never catch this, because it is not looking for it. This single blind spot causes more bad goals than every other criticism of SMART combined.

02

Achievable was not in the original, and it changed the framework

Doran wrote Assignable. Somewhere in the retelling it became Achievable, and a criterion about ownership turned into a criterion about ambition. Inside any system where goals carry consequences, “achievable” quietly teaches pre-negotiation: set what you can already do, then deliver it. The fix is not to abandon calibration. It is to label deliberate stretch and grade quarters on evidence, so honesty costs nothing.

03

It tests the sentence, not the choice

A goal can pass all five letters and still be the wrong thing to spend a quarter on, beautifully expressed. Nothing in SMART asks what you are choosing not to do, whether this is the pivotal problem, or whether anyone would notice if it were achieved.

04

Time-related invites calendar-shaped goals

The date in a good goal comes from the work, not from the end of the quarter. Read strictly, SMART is neutral on this. Read the way it is used, it produces a set of goals that all mysteriously finish on 31 December.

05

It is silent on how many

Twelve SMART goals is not a plan, it is a workload with punctuation. Every framework that came after SMART added the same missing thing in its own vocabulary: a limit, and a reason for the limit.

06

Doran’s own footnote, forgotten

He never intended all five criteria to apply to every goal at every level. Modern usage turned a set of considerations into a compliance checklist. The nuance he wrote is the same division of labour we use now: the Objective carries meaning and ambition, the Key Results carry the SMART discipline.

// smart vs okrs

Not rivals. Different altitudes.

SMART OKRs →
What it is A test applied to one goal sentence. A system for choosing and running a few goals.
Scope One goal at a time. A portfolio, with a deliberate limit on how many.
Ambition Realistic, and Achievable in the later variants. Stretch is legitimate, provided it is labelled as stretch.
Ownership Yes. Assignable was one of the original five. Yes, plus a team drawn around the goal rather than the org chart.
Says how many goals No. Nothing in SMART argues against twelve of them. Yes. One or two Objectives per team, two to four Key Results each.
Says why this goal No. It has no opinion on whether the goal is worth having. Yes: the narrative, the strategic choices, and the crux it targets.
Rhythm and feedback No. Weekly confidence, grading at the end of the cycle, and a retrospective.
Tells a task from an outcome No. This is the important one. Yes, and that distinction is most of the discipline.
Origin George Doran, Management Review, 1981. Andy Grove at Intel in the 1970s, named OKRs at Google from 1999.
Our verdict Keep it. Every good Key Result is a SMART sentence. Keep it. It decides which sentences are worth writing.

Every system on the goal-setting shelf adds the same missing layer in its own vocabulary. Hoshin Kanri adds direction and dialogue, EOS adds the weekly heartbeat, V2MOM adds narrative and obstacles, OKRs add alignment, cadence, confidence and grading. Pairing SMART sentences with a system around them is not a compromise. It is how the pieces were always meant to stack.

// asked and answered

Eight straight answers.

QWhat does SMART stand for?

In George Doran's original 1981 article: Specific, Measurable, Assignable, Realistic and Time-related. The version most people are taught today usually reads Specific, Measurable, Achievable, Relevant and Time-bound. The substitution matters: Assignable was about who owns the goal, while Achievable is about how ambitious it is allowed to be.

QWho invented SMART goals?

George T. Doran, in an article in Management Review in November 1981. He was writing about how managers could write clearer objectives, and he explicitly said he did not expect every goal at every level to satisfy all five criteria. That caveat is the part modern usage dropped.

QAre SMART goals still relevant?

Yes, at the sentence level, where they are excellent. SMART compresses the two most robust findings in goal-setting research, specificity and measurability for feedback, into five letters anyone can remember. What it does not do is decide which goals are worth having, how many you should run, or whether the thing you have written is an outcome or a task.

QWhat is the difference between SMART goals and OKRs?

SMART is a test you apply to one goal sentence. OKRs are a system for choosing a few goals and running them: which goals, why these ones, at what rhythm, with what feedback and grading. They are not rivals, because every well-written Key Result is already a SMART sentence. SMART writes a good sentence; a goal system decides which sentences are worth writing.

QCan a Key Result be a SMART goal?

It should be. A Key Result written as a movement from a baseline to a target by a date, with a named owner, satisfies SMART automatically. The reverse does not hold: a SMART goal is often a task, and a task can be completed in full while the outcome you wanted never moves.

QWhat is wrong with the A in SMART?

Two things. It was originally Assignable, meaning one named owner, and the substitution to Achievable lost that. And Achievable, read inside a company where goals affect pay, reputation or reviews, teaches people to commit to what they can already deliver. Keep calibration honest, label the goals where you are deliberately reaching, and grade on evidence rather than on the percentage.

QWhat is a good example of a SMART goal at work?

“Raise the share of new customers activated within 14 days from 38% to 60% by 30 September, owned by Sam.” It names the measure, the baseline, the target, the date and the owner. Note what makes it good is not the five letters: it is that the sentence describes a change in the world rather than work getting done.

QHow many SMART goals should a person or team have?

SMART itself has no answer, which is its main structural gap. In practice, one or two goals per team that genuinely matter, with everything else held steady on a scorecard of ongoing measures. A list of twelve equally weighted goals is a decision nobody made.

// connected concepts
Goal setting: the whole shelf →Goal-setting theory →What is a Key Result? →OKR examples →Task-based Key Results →Aspirational targets →Management by Objectives →OKR glossary, 101 terms → The full knowledge system →

SMART is credited to George T. Doran, Management Review, November 1981. The critique and the machinery connections are ZOKRI methodology. This is chapter 7 of From Trading to Scaling, “The Goals”. The book is free and takes about an hour.

// put it to work

Your teams already write SMART sentences. The question is which sentences are worth a quarter.

We install the system that chooses them, and the grading that keeps them honest.

Scale Ready →OKR coaching →Try the free AI coach →
A free working session
Want to pressure-test how you operate?

Ninety minutes with us, on your own or with your leadership team, then written recommendations five working days later. Free, and there is no pitch, proposal or follow-up.

See how it works
// the book and the install pack

Six things turn a company that trades into one that scales.

An hour to read, five working templates and the AI install pack. Ten years of strategy, operating systems, goal frameworks and culture building, battle tested in engagements that increased growth through better use of people, time and opportunity.

Get it for £100 →The only place the install pack is available.