Sales Goals: The Ones a Team Can Actually Act On

Sales Goals: The Ones a Team Can Actually Act On

A sales leader opens the planning document in the first week of January and writes four lines. Grow revenue by a third. Improve win rate. Shorten the sales cycle. Increase average deal size. Everybody in the room agrees with all four, because there is nothing in any of them to disagree with, and by March the only one anybody has looked at again is the first.

Those four lines are the standard output of sales goal setting, and they share one property: not one of them names anything a person could do differently on a Tuesday. They are descriptions of a preferred future written in the grammar of a target.

A goal that changes no behaviour is a wish with a number attached. Getting past that is the work, and it starts by separating three objects the word routinely blurs.

A goal, a quota and a KPI are three different objects

They arrive in the same meeting and get used interchangeably, which is why goal-setting exercises so often produce the revenue plan restated three ways.

A quota is an assignment. It is a number given to a person or a team for a period, against which their performance is judged, and it is a management instrument rather than a prediction. How it gets built, and the reconciliation step that gets skipped, is the subject of how a sales quota gets set and how it gets gamed.

A KPI is an instrument. It is a metric that has been promoted out of the hundreds a system can produce because somebody will act on it, and it needs a decision, a target and an owner attached before it earns the label. Which numbers change a decision works through the admission test.

A goal is a chosen intention for a period. It is the thing the team has decided to change, stated so that everybody can tell at the end whether it happened. It sits above the quota, which is how the goal gets distributed, and it is read through KPIs, which are how anybody knows where it stands.

Confusing them has one common outcome. The team sets the revenue number as its goal, distributes it as quota, reports it as a KPI, and has three artefacts saying the same thing and nothing at all saying what would have to change for it to happen.

The test that separates a goal from a wish

There is one question, and it is answerable in the meeting where the goal is proposed.

Name the thing somebody does differently next week if this goal is real.

The instruction to improve win rate fails it. Nobody does anything differently on Monday because of it, and at the end of the period the number will have moved for reasons nobody can attribute. "Disqualify on the criteria before an opportunity is created, and hold the loss reasons against a fixed list" passes it, because it names a behaviour, and win rate is then the number that says whether the behaviour worked.

That is the structural point underneath most goal-setting advice. Outcome numbers are how you find out. They are not what anybody works on, because nobody has direct access to them.

Outcome goalHow you find out

• Revenue against plan, win rate, average deal size, cycle length
• Nobody can act on it directly
• Moves a full sales cycle after the work that produced it
• Right for judging the period
• Fails as: the only line on the page

Mechanism goalWhat anybody works on

• A named behaviour, volume or standard somebody controls
• Actionable in the week it is set
• Moves within days, and is visible while there is still time
• Right for running the period
• Fails as: an activity count that becomes the thing being rewarded

The second column carries its own risk and it is worth naming early. Any mechanism goal expressed as a raw count will be satisfied as a raw count. A goal of logged contact attempts produces logged contact attempts. The defence is to express mechanism goals as ratios or as standards wherever the arithmetic allows, so that satisfying the goal requires the thing you actually wanted.

Where the number comes from

A goal with no stated origin cannot be argued with, and a goal nobody can argue with is one nobody has agreed to.

Two origins survive scrutiny. The first is your own history read on a stable definition: better than the same team, measured the same way, last period. The second is derivation from the plan, where the required volume at each stage falls out of the arithmetic rather than out of a comparison.

Derivation is the more useful of the two, because it produces a finding rather than only a target. The chain runs in one direction and every step uses a figure the business already holds.

1. Step 1Start from the plan

   The new revenue the period has to produce, separated from renewals and expansion

2. Step 2Divide by deal value

   Median rather than mean, so one outsized deal in the history does not set everybody's number

3. Step 3Apply win rate

   The historical rate at which qualified opportunities close, on a definition written down

4. Step 4Apply meeting conversion

   The rate at which held meetings become qualified opportunities

5. Step 5Read the required volume

   The number of qualified conversations the period needs, which is now a goal somebody can work on

Worked through with invented inputs, the shape becomes obvious. Suppose the plan asks for twelve new customers, the median deal has historically closed at one in four qualified opportunities, and one held meeting in three becomes a qualified opportunity. Every figure in this paragraph is invented for the illustration and describes no real team. That chain requires forty-eight qualified opportunities and a hundred and forty-four held meetings, which is a little over twelve meetings a month, every month, before anybody has allowed for seasonality or ramp.

The value of doing it is rarely the target at the end. It is the moment somebody notices that the plan requires roughly double last year's meeting volume with the same headcount, which is a solvable problem while it is still a sentence and an unsolvable one when it arrives as a missed quarter. The same reconciliation logic, applied to the assignment rather than to the goal, is in the bottom-up build described in the quota piece, and the cost side of the same plan sits in how a sales budget differs from a forecast.

What does not survive scrutiny is a benchmark table. Published averages aggregate companies with different segments, contract sizes, cycle lengths and definitions, and the definitional variation between them is normally wider than the difference the table claims to show. A target lifted from one is a target nobody in the room can defend.

SMART, and the letter it leaves out

The framework almost every goal-setting guide reaches for makes a goal writable. It does not make it achievable, and the gap between those two is where the exercise usually goes wrong.

Specific, measurable and time-bound are genuine improvements, and most vague goals fail on all three at once. Relevant is a filter worth running. The letter that carries the weight is achievable, and the framework offers no method for settling it, which is a problem because achievable is the only word in the acronym anybody actually argues about.

The derivation above is the method the acronym is missing. A goal is achievable when the chain from the plan back to the required volume closes using rates the team has actually produced. When it does not close, the honest responses are to add capacity, change the target list, improve one rate deliberately, or change the plan. Dividing the shortfall across the team instead converts a structural finding into a personal failure, one period at a time.

Does this goal earn its place

• ✓Yes: Somebody can name what they do differently next week because of it
• ✓Yes: Its origin is stated: your own history, or the arithmetic from the plan
• ✓Yes: One named person owns it
• ✓Yes: The definition, including the denominator, is written down and dated
• ✓Yes: It can move within the interval at which it will be read
• ✕No: Its number came from a published benchmark table
• ✕No: It is the revenue plan restated in different words
• ~Depends: It is a raw activity count that decides pay

Cascading, and the part that gets divided away

A company goal divided by headcount is arithmetic. It becomes a goal for a team only when the mechanism travels with the number.

The version that works keeps two things attached at every level. The outcome the level is accountable for, and the mechanism that level controls. A regional leader's mechanism is coverage and capacity. A team lead's mechanism is qualification standards and pipeline hygiene. An individual seller's mechanism is the conversations they create and the discipline they hold in them. Dividing the outcome downward without the mechanism produces four levels all working on the same lagging number and none of them working on anything.

The second requirement of the cascade is subtraction. Three goals per level is generally the ceiling, because attention is the scarce resource being allocated and a list of nine is a list of none. Adding a fifth goal should require removing one, which is the only rule that reliably keeps a plan readable.

The goal most teams never set

Sales goal lists are dominated by conversion and efficiency. Improve win rate, raise deal size, shorten the cycle, reduce churn. Every one of them is a goal about what happens to conversations after they exist.

Supply is the goal that goes unwritten, and it is frequently the binding constraint. A team with an excellent win rate and half the required meeting volume will miss, and no amount of work on the first number closes the gap. Pipeline coverage is where this usually surfaces, and it surfaces late, because a coverage ratio is a claim about a win rate as much as it is a claim about pipeline.

Setting a supply goal has one requirement that stops most teams from doing it: qualified has to mean something written down before the period starts. A goal of qualified conversations with no agreed definition of qualified is a goal that will be met by loosening the definition. Our own commercial standard is the same instrument, which is that meetings are qualified against criteria agreed in writing before launch, so the definition cannot be read off the result afterwards.

Three non-revenue goals are worth more than most of the standard list, because each names a decision:

Loss-reason concentration. Reasons collected from a fixed list, not free text, and reviewed for concentration. Reasons concentrating on seniority or fit are an instruction about targeting. Reasons concentrating on timing are usually an instruction about the segment.

Ramp spread. The gap between the fastest and slowest new seller reaching a defined competence bar, rather than the average, which is dominated by hiring.

Definitional stability. The count of stage and qualification definitions changed mid-period, with the target being zero. Attainment moves when the ruler moves, and nothing in the market has to happen for it.

Reading them at the interval they can move

A goal read weekly that can only move quarterly produces noise and a standing invitation to explain variance that is not there. Read mechanism goals weekly, outcome goals per period, and diagnostics when a gap needs explaining rather than on a schedule.

The related discipline is deciding in advance what a miss means. One seller in twelve short of the number is plausibly about that seller. Ten of twelve short is a statement about the number, and treating it as twelve performance conversations is a decision to keep the plan and change the people. Quota attainment is the measurement that makes that distinction visible, and the wider set of numbers that measure the machine rather than the person is in six sales operations KPIs.

The short version

Separate the three objects. A goal is a chosen intention, a quota is how it gets assigned, and a KPI is how anybody reads where it stands. Setting the revenue plan as all three produces three artefacts saying one thing.

Test every proposed goal by naming what somebody does differently next week because of it. Pair each outcome goal with a mechanism goal somebody controls, and express mechanisms as ratios or standards rather than as raw counts, because a raw count attached to pay will be satisfied literally.

Derive the number from the plan or from your own history on a stable definition, never from a benchmark table, and treat a chain that does not close as the finding rather than as an error to divide away. Cap the list at three per level, carry the mechanism down the cascade with the number, and set the supply goal that standard lists leave out.

If the arithmetic says the constraint is the number of qualified conversations rather than what happens inside them, that is the half we run: see what a first campaign produces.