How to Define KPIs for PreSales, Sales and Post-Sales Teams: A Revenue-Function Metric Guide

Key Takeaways

• PreSales, Sales and Post-Sales are three different jobs with three different failure modes. One shared dashboard measures none of them well.
• Split every function's metrics into controllable inputs, owned outcomes and shared outcomes. Compensating on an outcome a person cannot move produces gaming, not performance.
• The handoffs between the three functions leak more value than any single function's execution. Instrument the handoffs with their own metrics, owned jointly.
• PreSales is measured on technical win rate and solution fit, not on activity. A solution engineer judged on demo count will run demos that should never have been scheduled.
• Sales is measured on qualified pipeline coverage, stage conversion, win rate, cycle length and quota attainment — and quota attainment alone is a lagging number that tells you nothing about why.
• Post-Sales is measured on time-to-value, gross retention, net revenue retention and expansion. Counting tickets closed measures activity, not the outcome the customer bought.
• A KPI you cannot define the exact numerator and denominator for is not a KPI yet. Ambiguous definitions are where dashboards quietly diverge from reality.

1. The Short Answer: How to Define Revenue-Team KPIs

Define each function's KPIs by what that function actually controls. PreSales owns technical win rate, solution fit and time-to-technical-validation. Sales owns qualified pipeline coverage, stage-to-stage conversion, win rate, sales-cycle length and quota attainment. Post-Sales owns onboarding time-to-value, gross revenue retention, net revenue retention and expansion. Then add a small number of handoff metrics — owned jointly — because the seams between the three functions leak more value than any single function's execution.

The mistake this guide exists to prevent is measuring three different jobs with one dashboard. PreSales, Sales and Post-Sales fail for different reasons, on different timescales, against different definitions of success. A revenue leader who watches only bookings sees the symptom of all three and the cause of none.

This is the metric layer of the same discipline we apply in our [revenue operations](/revenue-operations) practice: define the outcome, define what moves it, and measure the two separately so the number tells you what to do rather than only how you did.

• AEO Quick Answer: Measure each function on the outcome it controls, split controllable inputs from owned outcomes, and add joint handoff metrics between functions.
• Three functions, three failure modes, three definitions of success — not one dashboard.
• The largest leak is almost always at a handoff, not inside a function.

2. Why One Revenue Dashboard Measures Nothing Well

Most revenue orgs run on bookings, pipeline and a churn number. Those are real, and they are all lagging outcomes that aggregate the work of three functions into figures no single team can act on.

Consider a quarter where bookings miss. Was the problem that PreSales validated deals that were never technically winnable, so they consumed cycle time and died late? That Sales built insufficient pipeline three months earlier? That Post-Sales let last year's cohort churn, so net revenue went backwards before a single new deal closed? The bookings number is identical in all three cases and points at none of them.

The three functions also operate on different clocks. PreSales effects show up within a sales cycle. Sales pipeline effects show up one to two cycles later. Post-Sales retention effects show up over quarters and years. Averaging them onto one monthly dashboard blends signals that move at completely different speeds, which is how a retention problem gets misdiagnosed as a sales problem and answered with a sales-hiring plan.

And they fail for structurally different reasons. PreSales fails on technical fit and credibility. Sales fails on qualification, multithreading and urgency. Post-Sales fails on time-to-value and ongoing outcome delivery. A metric that surfaces one of these is usually blind to the other two.

The fix is not more dashboards. It is the right small set per function, each measuring an outcome that function can move, plus explicit handoff metrics so the seams are visible.

• Bookings, pipeline and churn are lagging aggregates no single team can act on.
• The three functions resolve on different clocks — a cycle, two cycles, several quarters.
• They fail for different reasons, so one metric is blind to two of the three.

3. The Framework: Controllable Inputs, Owned Outcomes, Shared Outcomes

Before naming a single metric, classify it. Every candidate KPI falls into one of three tiers, and the tier decides how you use it — especially whether you compensate on it.

Controllable inputs are things a person can do directly, today, regardless of anyone else. Number of technical validations completed. Multithreaded contacts added to a deal. Onboarding calls held. These are useful for coaching and for diagnosing effort, and they are dangerous as compensation targets because they are trivially gameable — you can always do more of an activity without doing it well.

Owned outcomes are results a person or team materially controls but does not do mechanically. A solution engineer's technical win rate. A rep's stage-two-to-three conversion. A CSM's onboarding time-to-value. These are the right basis for evaluation, because they reward the outcome rather than the motion, while still being close enough to the person's actions to be fair.

Shared outcomes are results multiple functions jointly determine. Overall win rate depends on lead quality, technical fit and closing skill. Net revenue retention depends on the product, onboarding, ongoing success and the quality of what Sales sold in the first place. These belong on the leadership dashboard and in shared incentives, never as the sole measure of one function, because holding one team solely accountable for an outcome three teams determine guarantees blame rather than improvement.

The single most common compensation error in revenue orgs is paying an individual on a shared outcome. It feels rigorous — you are paying for results — and it reliably produces resentment and gaming, because the person is being judged on things outside their control. Pay on owned outcomes; watch shared outcomes at the team level. This is the same accountability logic behind our [quality scoring](/solutions/quality-scoring) work.

• Controllable inputs: coaching and diagnosis only, never compensation — they are gameable.
• Owned outcomes: the correct basis for individual evaluation and pay.
• Shared outcomes: leadership dashboard and team incentives, never one function alone.
• The classic error: paying an individual on an outcome three functions jointly determine.

4. PreSales KPIs: Measuring Technical Credibility and Fit

PreSales — solution engineering, solution consulting, sales engineering — exists to establish that your product actually solves the buyer's technical problem, and to do it credibly enough that the technical evaluators become advocates rather than blockers. Its metrics must measure that, not activity.

Technical win rate. The owned outcome that matters most: of the opportunities PreSales engaged and technically validated, what proportion were won. This isolates the PreSales contribution better than overall win rate, because it conditions on deals they actually touched. A falling technical win rate points at either solution fit or the quality of the validation, both of which PreSales influences directly.

Solution-fit rate, or qualified-out rate. The proportion of engaged opportunities that PreSales correctly identifies as a poor technical fit and helps disqualify early. This is counterintuitive and important: a good PreSales function kills bad-fit deals fast, before they consume a full sales cycle. Measuring only wins punishes exactly the disqualification that protects the pipeline's health. A healthy function has a visible, non-zero qualified-out rate.

Time-to-technical-validation. The elapsed time from PreSales engagement to a documented technical win — proof of concept passed, security review cleared, architecture accepted. Shorter is better up to a point, and the distribution matters more than the average: a long tail of validations that never resolve is usually deals that should have been disqualified.

Proof-of-concept success rate. Where POCs or trials are part of the motion, the proportion that meet their pre-agreed success criteria. This only means something if the success criteria were defined before the POC started — a POC with no exit criteria is unmeasurable and usually unwinnable, the same problem as a sales stage with no exit criterion, which we cover in the [funnel stages guide](/guides/funnel-stages-guide).

What not to measure PreSales on: demo count, hours logged, or deals touched. A solution engineer judged on demo volume will run demos that should never have been scheduled, which lengthens cycles and lowers technical win rate — the metric actively works against the outcome. Activity is a coaching input, never a PreSales KPI.

The PreSales-specific handoff metric: technical-validation-to-close rate, owned jointly with Sales. If deals validate technically but then stall, the problem is commercial, not technical, and this metric makes that visible instead of leaving the two functions to blame each other.

• Technical win rate — the core owned outcome, conditioned on deals PreSales engaged.
• Solution-fit / qualified-out rate — rewards fast disqualification of bad-fit deals.
• Time-to-technical-validation — watch the tail, not just the median.
• POC success rate — meaningful only with pre-agreed exit criteria.
• Never: demo count or hours — activity metrics that work against the outcome.

5. Sales KPIs: The Metrics That Actually Diagnose

Sales is the most over-measured function in most companies and often the least well measured, because the dashboard is crowded with activity counts and one lagging outcome — quota attainment — with little in between that explains the outcome.

Quota attainment. The lagging outcome everyone watches: bookings against target. It is necessary and nearly useless as a diagnostic, because it tells you the result and nothing about the cause. Track it, but never stop there. See the [quota attainment](/glossary/quota-attainment) definition for the mechanics.

Qualified pipeline coverage. The ratio of qualified open pipeline to the target for the period, usually expressed as a multiple. This is the most predictive leading indicator Sales has, because it exposes a shortfall months before it becomes a bookings miss. The right coverage multiple is derived from your own win rate and cycle length, not from a rule of thumb — a team that wins 33% needs roughly 3x, a team that wins 20% needs 5x, and adopting someone else's multiple imports their win rate along with it.

Stage-to-stage conversion. The proportion of opportunities advancing between each pipeline stage, measured cohort-based. This is the diagnostic layer quota attainment lacks: it localises where deals are actually lost, which is the difference between a targeting problem, a qualification problem and a closing problem. It only works if stages have verifiable exit criteria.

Win rate. The proportion of qualified opportunities won, segmented by source, segment and product. A blended win rate hides the variance that matters — the correct action for a low win rate on one source is entirely different from a low win rate across the board.

Sales-cycle length. The median and the distribution of time from qualified opportunity to close. Rising cycle length is an early warning of deteriorating qualification or growing deal complexity, and the distribution shape tells you whether you are running one motion or two.

Average deal size and its trend, because a quota hit on fewer larger deals is a different risk profile from the same quota on many small ones.

Multithreading depth — the number of engaged contacts per opportunity, with roles. Single-threaded deals close at materially lower rates, and this input predicts win rate well enough to coach on. It is a controllable input, so coach on it; do not compensate on it.

The controllable inputs — calls, emails, meetings booked — belong in coaching conversations and nowhere near the compensation plan. A rep paid on activity produces activity, and activity is not revenue.

• Quota attainment — the lagging outcome; necessary, but never a diagnosis on its own.
• Qualified pipeline coverage — the most predictive leading indicator; derive the multiple from your own win rate.
• Stage-to-stage conversion — localises where deals are lost.
• Win rate and cycle length — always segmented; the distribution matters.
• Multithreading depth — coach on it as an input, never pay on it.

6. Post-Sales KPIs: Measuring Retained and Expanded Value

Post-Sales — customer success, account management, onboarding, support — determines the majority of lifetime value in any recurring-revenue business, and it is the function most often measured on activity (tickets closed, calls made) rather than on the outcome the customer actually bought.

Onboarding time-to-value. The elapsed time from contract signature to the customer achieving their first defined outcome. This is the single most predictive early metric in Post-Sales, because time-to-value in the first weeks correlates strongly with long-term retention. It requires defining 'value' concretely per product — a milestone a second person could verify — not a vague sense that onboarding went well.

Gross revenue retention. The proportion of recurring revenue retained before any expansion, over a period. This isolates pure churn and downgrade, and it cannot exceed 100% — which is exactly why it is honest. A high gross retention number is hard to fake with upsells to a few large accounts.

Net revenue retention. Recurring revenue retained including expansion, so it can exceed 100%. The [NRR](/glossary/nrr) figure is the headline health metric for a subscription business, but it must be read alongside gross retention, because strong expansion in a few accounts can mask serious churn in the base.

Logo retention, tracked separately from revenue retention, because losing many small customers and keeping a few large ones can look healthy on revenue while the market position erodes.

Expansion rate — the proportion of the base that grew, and the revenue from it. This is the owned outcome for account management specifically, distinct from the CSM's retention focus.

Product adoption depth — the proportion of purchased capability actually in use. Low adoption is the leading indicator of churn that has not happened yet, which makes it more actionable than the churn itself. This is where a customer health score earns its place, if it is built from behaviour rather than from CSM sentiment.

Net promoter or satisfaction, treated as a directional input rather than an owned outcome, because it is influenced by many things outside Post-Sales and is easily gamed by asking at favourable moments. Our [NPS modelling](/solutions/nps-modeling) work is about making that signal reliable enough to act on.

What not to measure Post-Sales on: tickets closed, response time alone, or call volume. These measure motion. A CSM optimising ticket closure will close tickets, not retain customers, and the two are not the same thing.

• Onboarding time-to-value — the most predictive early Post-Sales metric; define 'value' concretely.
• Gross revenue retention — honest because it cannot exceed 100%.
• Net revenue retention — the headline, read alongside gross so expansion cannot mask churn.
• Adoption depth — the leading indicator of churn before it happens.
• Never: tickets closed or call volume — motion, not the retained outcome.

7. The Handoff Metrics: Where the Real Leaks Live

The most valuable metrics in a revenue org sit between the functions, not inside them, because that is where accountability is ambiguous and value leaks quietly.

Sales-to-PreSales engagement quality. Of the deals Sales brings to PreSales, what proportion are genuinely qualified for technical engagement. A low figure means Sales is using PreSales as a crutch to progress unqualified deals, which burns the scarcest and most expensive resource in the revenue org on deals that will not close.

PreSales-to-close rate. Of technically validated deals, what proportion close. A high technical win rate with a low close rate means deals are winnable and not being won — a commercial problem that would otherwise be invisible, with each function blaming the other.

Sales-to-onboarding fidelity. Whether what Sales sold matches what Post-Sales can deliver. Measured through early-churn rate and through the frequency of expectation mismatches surfaced in onboarding. A spike here means Sales is overselling to hit quota, borrowing from next year's retention to pay for this quarter's bookings — the most expensive and most common cross-function failure there is.

Expansion-signal-to-Sales latency, where expansion is sold by Sales rather than by account management: how quickly a Post-Sales-detected expansion signal reaches someone who can act on it. Slow latency here is pure lost revenue on demand you already created.

These metrics must be owned jointly, appear on both functions' dashboards, and feed a shared incentive component. A handoff metric owned by neither function is a handoff metric nobody fixes. Instrumenting these is usually the highest-return measurement work a revenue org can do, and it is central to our [business operations](/solutions/business-ops) engagements.

• Sales-to-PreSales engagement quality — is PreSales being spent on qualified deals?
• PreSales-to-close rate — winnable deals that are not being won are a commercial problem.
• Sales-to-onboarding fidelity — early churn exposes overselling to hit quota.
• Expansion-signal latency — slow routing of expansion signals is lost revenue on existing demand.
• Handoff metrics must be jointly owned, or nobody fixes them.

8. Leading Versus Lagging, and Why You Need Both

Every metric is either leading — it moves before the outcome — or lagging — it reports the outcome after the fact. A dashboard of only lagging metrics tells you what happened when it is too late to change it; a dashboard of only leading metrics is a pile of activity with no confirmation it produced anything.

The pattern that works pairs each lagging outcome with the leading indicators that predict it. Quota attainment (lagging) is predicted by qualified pipeline coverage and stage conversion (leading). Net revenue retention (lagging) is predicted by onboarding time-to-value and adoption depth (leading). Technical win rate (lagging within a cycle) is predicted by solution-fit rate and time-to-validation.

Manage forward on the leading indicators and confirm on the lagging ones. If pipeline coverage is healthy and conversion holds but bookings still miss, something changed that your leading indicators do not capture, and you have found a gap in your measurement rather than a performance problem. That is a useful discovery, not a failure.

The discipline that makes this real is validating your leading indicators against the lagging outcomes periodically. A leading indicator that stops predicting its outcome has drifted and needs recalibration — pipeline coverage assumptions built on last year's win rate mislead once the win rate moves.

• Pair every lagging outcome with the leading indicators that predict it.
• Manage forward on leading, confirm on lagging.
• Healthy leading indicators plus a missed outcome means a measurement gap, not just underperformance.
• Revalidate that your leading indicators still predict their outcomes.

9. Defining a KPI Precisely: Numerator, Denominator, Window

A metric name is not a KPI. 'Win rate' is a name; a KPI is the exact numerator, the exact denominator, the time window and the segmentation. Two teams computing 'win rate' three different ways will argue about the number forever, and both will be partly right.

For every KPI, write down four things. The numerator: exactly what counts as a success — closed-won signed, or verbally agreed. The denominator: exactly the population — all opportunities, or only qualified ones, and does a deal that is still open count. The window: cohort-based, following the deals that entered in a period to their outcome, or a period ratio of two totals. And the segmentation: by source, segment, product, and owner, because the blended number hides the variance that determines the action.

The window choice matters more than teams expect. A period ratio — deals won this quarter over deals closed this quarter — distorts badly whenever volume is changing or cycles are long, because the wins and losses in the numerator and denominator came from different cohorts. Cohort-based measurement, following a set of deals from entry to outcome, is almost always more honest, and it is the only way to compute conversion that means anything over a long sales cycle.

Write these definitions down in one place, agreed across functions, and treat changing a definition as a versioned decision rather than a quiet spreadsheet edit. A KPI whose definition drifts is worse than no KPI, because people trust it while it lies. This is the same rigour we bring to canonical definitions everywhere in a revenue system.

• A KPI is a numerator, a denominator, a window and a segmentation — not a name.
• Cohort-based windows beat period ratios whenever volume moves or cycles are long.
• Segment by source, segment, product and owner as a default.
• Version definition changes; a silently drifting KPI lies while being trusted.

10. How Many KPIs Per Role, and Who Sees What

More metrics is not more insight. A role with fifteen KPIs has no priorities, because when everything is measured nothing is emphasised. The workable pattern is a small number of owned KPIs per role plus visibility into the team's shared outcomes.

For an individual contributor: three to five owned KPIs, of which one or two are the primary evaluation basis and the rest are diagnostic. A rep's primary is quota attainment against a fair quota; their diagnostics are pipeline coverage, conversion and cycle length. A CSM's primary is net revenue retention on their book; their diagnostics are time-to-value and adoption depth.

For a team lead: the aggregate of their team's owned outcomes plus the handoff metrics they share with adjacent functions. This is where shared outcomes belong — a sales manager can and should be accountable for the team's win rate in a way no individual rep can be.

For the revenue leader: the lagging outcomes across all three functions, the handoff metrics, and the small set of leading indicators that predict next quarter. This dashboard should fit on one screen and answer one question — where is the constraint right now — rather than displaying everything measurable.

The visibility principle: everyone sees the team's shared outcomes, because transparency on the collective result builds cooperation, while individual owned metrics are between the person and their manager. Publishing individual rankings company-wide reliably produces gaming and sandbagging, not performance.

• Individual: three to five owned KPIs, one or two primary, the rest diagnostic.
• Team lead: aggregated team outcomes plus shared handoff metrics.
• Revenue leader: cross-function lagging outcomes, handoffs, and predictive leading indicators, on one screen.
• Shared outcomes are transparent; individual metrics stay between person and manager.

11. The Incentive Traps That Make KPIs Lie

Any measured number that is tied to reward creates pressure to satisfy the measure rather than the intent — Goodhart's law, and it is not avoidable, only manageable. The goal is to design metrics whose easiest path to a good number is also the path to the real outcome.

The activity trap. Paying on calls, demos or tickets produces calls, demos and tickets, decoupled from results. Fix: compensate on owned outcomes, use activity only for coaching.

The sandbagging trap. When quota is set from last period's performance, over-performing this period raises next period's target, so rational reps hide capacity. Fix: set quota from territory potential and market, not from a ratchet on individual history.

The overselling trap. When Sales is paid on bookings alone, the cheapest path to a booking is to promise more than the product delivers, which shows up as churn in Post-Sales two quarters later. Fix: a retention or early-churn clawback component, and the sales-to-onboarding fidelity metric on the shared dashboard.

The cherry-picking trap. When PreSales is measured on technical win rate alone, the safe move is to engage only easy deals and avoid hard ones. Fix: pair win rate with coverage of assigned deals, so avoidance is visible.

The vanity-retention trap. When Post-Sales is measured on net revenue retention alone, a few large expansions can hide a churning base. Fix: gross retention and logo retention alongside NRR.

The timing-game trap. Almost every period metric can be improved by moving an event across the period boundary — pulling a deal forward, delaying a churn. Fix: cohort-based measurement, which follows the deal regardless of when it is recorded, removes most of the incentive to game the calendar.

The general defence is the same each time: never compensate on a single number in isolation, pair every outcome with a guardrail metric that catches the gaming path, and measure cohorts rather than periods.

• Activity trap — pay on outcomes, coach on activity.
• Sandbagging trap — set quota from potential, not from a ratchet on history.
• Overselling trap — retention clawback plus sales-to-onboarding fidelity.
• Cherry-picking trap — pair win rate with coverage of assigned deals.
• Timing-game trap — cohort measurement removes most calendar gaming.

12. Building the System: A Practical Sequence

Implementing this is a few weeks of disciplined work, and it front-loads the definitional part that most teams skip.

Week one: map the three functions and their handoffs as they actually operate, including the exceptions. Identify, for each function, the one outcome it most controls. Resist adding metrics at this stage; you are finding the spine.

Week two: for each function, define the primary owned outcome and two or three diagnostic leading indicators, each with its exact numerator, denominator, window and segmentation. Write the definitions in one shared document and get cross-function sign-off, because the definitions are the substance.

Week three: define the handoff metrics and assign joint ownership. This is the step that distinguishes a revenue system from three siloed dashboards, and it is the one most likely to be resisted, because joint ownership is uncomfortable exactly where accountability was previously ambiguous.

Week four: instrument in the system of record. Decide where each metric is computed and confirm the CRM can produce cohort-based figures — if it cannot, that is the moment a warehouse becomes worth planning, and it is a input to your [CRM selection](/guides/choose-best-crm-guide). Build the three dashboards: individual, team, leadership.

Ongoing: review leading indicators weekly, outcomes monthly, and the metric definitions quarterly against whether the leading indicators still predict the lagging ones. Retire metrics nobody acts on — an unused metric is measurement overhead with no return.

• Week 1: map functions and handoffs; find the one outcome each function controls.
• Week 2: define primary and diagnostic metrics with exact numerator/denominator/window.
• Week 3: define and jointly assign handoff metrics — the step that makes it a system.
• Week 4: instrument, confirm cohort computation, build the three dashboards.

13. A Worked Example (Illustrative Model)

The figures below are an illustrative model to show how the layers connect. They are not client data.

A B2B software company misses its quarterly bookings target by 20%. The single bookings number offers no cause. The layered metrics do.

PreSales: technical win rate held at 68%, and qualified-out rate was healthy at 15%, so technical fit was not the problem — deals that engaged PreSales were being validated appropriately.

Sales: qualified pipeline coverage entering the quarter was 2.4x against a required 3.2x derived from the 31% win rate and the cycle length. The shortfall was visible a full quarter earlier as a coverage gap and was not acted on. Stage conversion and win rate were both stable, which rules out a closing or qualification problem.

Post-Sales: gross retention slipped from 91% to 86%, so net revenue went backwards before new bookings were even counted, deepening the apparent miss.

The diagnosis is now specific and actionable: a pipeline-generation shortfall one to two quarters earlier, compounded by a retention slip — not a closing problem, and not a PreSales problem. A team watching only bookings would likely have responded by pressuring the closers, which the conversion data shows would have changed nothing. The layered metrics point instead at demand generation and at the retention base, which is where the work actually is. And notice the leading indicator, pipeline coverage, flagged the bookings miss a quarter before it happened — which is the entire point of measuring forward.

14. Putting It Together

Defining revenue-team KPIs well is mostly an act of separation: separating three functions that fail for different reasons, separating controllable inputs from owned outcomes from shared outcomes, separating leading indicators from lagging ones, and separating the individual metrics used to coach from the shared metrics used to align.

Do that separation and the numbers start telling you what to do instead of only how you did. Skip it, and you are left with a bookings figure that moves for reasons you cannot see, on a timescale that makes it too late to act.

The highest-return single move for most revenue orgs is not adding metrics inside the functions — it is instrumenting the three handoffs, because that is where value leaks with nobody accountable. If your functions each look healthy but the aggregate disappoints, look at the seams first.

If you want the metric layer designed against your actual funnel and handoffs rather than against a template, that diagnostic is where our [revenue operations](/solutions/business-ops) engagements begin — and the parent framework for cascading these into departmental and role-level targets is our [KPI and KRA guide](/guides/set-kpis-kras-department-role-guide).