Formula reference · Account-free
Sales KPI Formulas: 20 Reproducible Metric Contracts
Twenty Sales and RevOps metric contracts with formulas, populations, periods, units, invalid states, and decision boundaries.
- Method owner
- Anastasiia Krynytska
- Reviewed
- 10 Sep 2026
- Reading
- 10 min read
Practical answer
What this resource does
A KPI is reproducible only when the formula travels with its business question, population, period, units, source fields, owner, and missing-data rule.
- No lead gate. Open, calculate, or download directly.
- Visible method. Definitions, formulas, ownership, and invalid states stay inspectable.
- Human authority. The asset supports a decision; it never owns one.
A sales formula is reproducible only when the population, events, period, units, source, exclusions, and error behavior travel with the number. This reference provides 20 metric contracts for revenue teams. Each contract states the formula, required fields, valid use, common failure, and invalid state. Copy the contract into a dashboard specification before writing a query.
These definitions follow the Luck My Sales operating method owned by Anastasiia Krynytska. Common alternatives are shown where they materially change meaning. Salesforce’s explanations of sales velocity and quota attainment and Stripe’s CAC guidance provide external category references. House definitions and targets are labeled; they are not universal benchmarks.
01 / Method
Before calculating: the metric-contract header
Store this header with every metric:
- metric name and version;
- business question and decision owner;
- numerator and denominator event definitions;
- eligible population and exclusions;
- time window, as-of timestamp, timezone, and cohort logic;
- currency, value basis, and conversion rule;
- source systems, stable keys, and authoritative fields;
- missing, duplicate, reopened, and late-arriving record treatment;
- display rounding and invalid-state behavior;
- owner, review date, and change history.
A dashboard label is not a contract. If two teams use different definitions, keep both names specific rather than forcing false consistency.
02 / Method
1. Quota attainment
Formula:
credited achieved revenue ÷ assigned quota × 100%.Population: one seller, team, or territory with revenue and quota aligned to the same period, currency, product scope, and crediting policy. Use: compare realized credited performance with the assigned target. Failure: bookings, revenue, and cash are mixed, split credit is double counted, or a changed quota rewrites history. Invalid: assigned quota is zero or missing; return unavailable, not infinity. Salesforce’s quota-attainment overview supports the general relationship, while each company owns its credit policy.
03 / Method
2. Average deal value
Formula:
total comparable deal value ÷ number of comparable deals.Population: usually won deals in a declared cohort; specify annual contract value, total contract value, recurring revenue, or another basis. Use: describe mean transaction size and support capacity models. Failure: currencies, renewals, expansions, services, or enterprise and SMB motions are blended without disclosure. Invalid: zero comparable deals; return unavailable. Report count and consider median and concentration beside the mean.
04 / Method
3. Opportunity win rate
Formula:
won opportunities ÷ eligible decided opportunities × 100%.Population: a cohort with a declared start event and final won/lost rule. Use: quantify the share of resolved eligible opportunities won. Failure: open deals enter the denominator, stage-entry cohorts are mixed with close-period outcomes, or duplicates and no-decisions are treated inconsistently. Invalid: no decided opportunities. Revenue win rate (
won value ÷ decided value) is a different metric and must be named separately.05 / Method
4. Gross open pipeline
Formula:
sum of open opportunity value in the declared scope.Population: opportunities open at one as-of timestamp under explicit type, stage, owner, period, and currency rules. Use: measure the current value stock before qualification. Failure: duplicates, expired deals, mixed value bases, or out-of-period close dates are included silently. Invalid: source extract or currency rule unavailable. Gross pipeline is not qualified pipeline, weighted pipeline, or forecast.
07 / Method
6. Pipeline coverage
Luck My Sales formula:
MEDDPICC-qualified pipeline value ÷ aligned revenue target.Population: qualified opportunities capable of contributing to the same target period, team, product, and currency. Use: inspect whether the evidence-backed pipeline stock is large relative to target. Common alternative: all open pipeline divided by quota. Label the alternative gross pipeline coverage. Failure: gross and qualified values are conflated, or next-quarter deals support this-quarter target. Invalid: target is zero or missing. Coverage is not forecast.
08 / Method
7. Required pipeline
Formula:
revenue target ÷ aligned historical win rate.Population: target and win rate for comparable segment, stage boundary, value basis, and period. Use: create a planning estimate under a simple conversion assumption. Failure: a blended or stale win rate is treated as a constant; capacity and cycle timing are ignored. Invalid: win rate is zero, missing, or based on an incompatible cohort. Display assumptions and scenarios, not false precision.
09 / Method
8. Stage conversion rate
Formula:
records reaching destination stage ÷ eligible records entering origin stage × 100%.Population: one entry cohort with a defined observation window and durable stage events. Use: find progression constraints between two states. Failure: current-stage snapshots replace event history, skipped or reversed stages are mishandled, or the cohort has not matured. Invalid: no eligible origin entries. State whether conversion is eventual or within a time limit.
10 / Method
9. Average sales-cycle length
Formula:
sum(end timestamp − start timestamp) ÷ eligible won deals.Population: comparable wins with declared start and end events. Use: describe mean elapsed time and feed capacity or velocity models. Failure: lead creation and opportunity creation are mixed, business and calendar days are confused, or only fast wins survive the filter. Invalid: no eligible wins or missing timestamps. Report median, distribution, and open aging when material.
11 / Method
10. Sales velocity
Formula:
eligible opportunities × average deal value × win rate ÷ average sales-cycle length.Unit: currency per declared time unit. Use: inspect the simplified relationship between volume, value, conversion, and time. Failure: inputs come from different populations or the output is called a forecast. Invalid: cycle length is zero or negative; required inputs are missing; currency/value bases conflict. Use the browser-local sales velocity calculator for substituted formulas and sensitivity.
12 / Method
11. Close-date slippage rate
Formula:
eligible opportunities whose expected close moved beyond the target boundary ÷ eligible opportunities expected to close in the original period × 100%.Population: a frozen starting snapshot joined by stable opportunity ID to later states. Use: measure timing instability. Failure: overwritten close dates erase history, already-won/lost records are mishandled, or administrative changes count as new pipeline. Invalid: no eligible starting opportunities. Report slipped value as a companion metric because count and revenue exposure differ.
13 / Method
12. Forecast accuracy
One transparent form:
1 − absolute(forecast − actual) ÷ absolute(actual), displayed with a declared floor or as forecast error where actual is zero.Population: a forecast frozen at a named time and the corresponding final actual under the same revenue definition. Use: quantify error magnitude for comparable forecast horizons. Failure: forecasts are overwritten, favorable over- and under-forecast errors cancel, or zero actuals create undefined ratios. Invalid: actual equals zero for this percentage form; report absolute error and signed error instead. Never cap or transform without disclosure.
14 / Method
13. Accepted-contact rate
Luck My Sales formula:
Accepted Contacts ÷ eligible contacted people × 100%.House definition: an Accepted Contact passes hard-bounce validation and ICP fit, then provides a conscious, two-way positive signal such as a reply or request. Use: measure dialogue-capable response from a declared outreach population. Common alternative: any enriched, sequenced, delivered, or clicked contact. Failure: automated opens or inferred intent are counted as conscious response. Invalid: no eligible contacted people. Preserve delivery, reply, positive reply, and acceptance as separate events.
15 / Method
14. Speed-to-Lead
Luck My Sales formula:
live-contact-capable timestamp − eligible first-party lead event timestamp.Output: median and percentile duration, plus share within a threshold. The house target is under 30 seconds for the defined workflow. House boundary: contact means a live or voice-dialogue-capable interaction; an automated email is not contact. Common alternative: time to any automated response or task creation. Failure: offline hours, duplicates, test leads, and routing failures disappear from the population. Invalid: either event timestamp is missing. Label the house target; do not state it as a universal benchmark.
16 / Method
15. Hard-bounce rate
Formula:
hard-bounced attempted messages ÷ eligible attempted messages × 100%.Population: one sending path, period, and provider taxonomy, retaining suppressed and rejected cases according to the declared rule. Use: monitor address quality and infrastructure outcomes. The owner specifies a Luck My Sales target below 0.5%. Failure: the denominator uses delivered messages, hiding bounces, or soft and hard categories are mixed. Invalid: zero attempted messages or missing event reconciliation. Email-provider definitions can differ; document mappings.
17 / Method
16. Enrichment match rate
Formula:
records with an accepted matched entity ÷ eligible submitted records × 100%.Population: records submitted under one identity and confidence policy. Use: measure coverage of accepted matches. Failure: any returned value counts as correct, duplicates inflate the numerator, or ambiguous entities are forced into a match. Invalid: no eligible submissions. Pair match rate with verified precision, provenance coverage, staleness, and correction rate; match volume alone does not prove data quality.
18 / Method
17. Customer acquisition cost
Luck My Sales formula:
fully loaded Sales and Marketing expense ÷ new customers acquired in the same period.Population: cost pool and new-customer denominator aligned by period, segment, geography, channel policy, and currency. Use: estimate average acquisition cost. Failure: media-only spend is labeled fully loaded CAC, or leads replace customers. Invalid: zero new customers or missing cost pool. Use the CAC calculator to expose components and validation.
19 / Method
18. Steady-state contribution LTV
Owner-specified simple formula:
ARPU × gross margin ÷ net churn.Population: recurring customer economics with ARPU and positive net churn expressed in the same period. Use: rough steady-state comparison under stable assumptions. Failure: revenue and logo churn are mixed, service cost is excluded without disclosure, or very low churn creates implausible precision. Invalid: net churn is zero or negative; return unavailable and use cohort contribution LTV. Never display infinity or a negative LTV from this formula.
Cohort contribution LTV instead sums realized and supportable forecast gross-profit contribution for a defined acquisition cohort over a declared horizon. It requires retention, expansion, contraction, service cost, timing, and forecast assumptions.
20 / Method
19. LTV:CAC
Formula:
valid LTV estimate ÷ aligned CAC.Population: the same or demonstrably compatible customer segment, acquisition model, currency, and margin policy. Use: compare lifetime contribution estimate with acquisition cost. Failure: enterprise LTV is divided by blended CAC, or invalid steady-state LTV is allowed through. Invalid: CAC is zero/negative under the chosen interpretation, or either component is unavailable. Owner-supplied context bands are at least 3:1 and approximately 4:1 for mature SaaS; label them as context, not universal verdicts.
21 / Method
20. CAC payback period
Simple formula:
CAC ÷ monthly gross-profit contribution per new customer.With stable recurring revenue, the denominator can be
monthly ARPU × gross margin. Use: estimate months needed for recurring gross-profit contribution to recover acquisition cost. Failure: annual and monthly units are mixed, implementation timing and service cost are ignored, or booked contract value replaces contribution. Invalid: contribution is zero or negative; return unavailable. Owner-supplied context bands are at or below 12 months generally and 8 months for SMB, but interpretation remains model-specific.22 / Method
How to implement the contracts in a warehouse
Preserve immutable raw events and build versioned semantic definitions on top. Stable entity IDs should survive name and owner changes. Store original timestamps, ingestion timestamps, source, actor, and corrections. A current-state table alone cannot reproduce stage conversion, slippage, response time, or forecast accuracy.
Separate measurement time from business-effective time. Late-arriving events may restate a historical metric; the system should record whether dashboards restate history or freeze published periods. Keep currency conversion rates, quota versions, ownership history, and territory mappings effective-dated.
Tests should cover normal rows, duplicates, missing IDs, zero denominators, negative corrections, reopened records, boundary timestamps, currency conflicts, and scope changes. Reconcile aggregates to authoritative finance or CRM outputs where appropriate. A successful query run is not proof that the business definition is correct.
23 / Method
Presentation rules
Display the population, period, unit, and last refresh beside the value. Show numerator and denominator counts where possible. Use Unavailable or Unknown with a reason instead of coerced zero. Distinguish observed data, derived values, forecasts, and targets visually and semantically.
Do not add red/amber/green labels unless an accountable owner has defined thresholds, evidence, review cadence, and action. A target is not an empirical benchmark. A benchmark is not a causal recommendation. A metric is not a decision.
24 / Method
Frequently asked questions
Why do some formulas differ across tools?
Tools optimize for different objects, stages, and reporting models. Use the definition that matches the business question, name it precisely, and preserve alternatives when comparison requires them.
Should zero and blank be treated the same?
No. Zero is an observed or asserted value; blank is missing. Their operational meaning and error behavior differ.
How often should contracts be reviewed?
At a scheduled cadence and whenever systems, fields, stages, pricing, quota, crediting, or business motion changes. Version material changes and mark discontinuities.
Are the house thresholds industry standards?
No. They are owner-specified operating targets or context bands. They are labeled so teams can adapt them without mistaking them for universal evidence.
Method owner: Anastasiia Krynytska, RevOps Architecture Lead. Reviewed 10 September 2026. This reference supports metric design; it is not accounting or financial advice.
Method & privacy note
Inspect before you act
- The complete approved editorial text is published on this page, not hidden behind the download or interface.
- Calculator inputs stay in browser memory; workbook contents stay wherever the user saves them.
- Unknown, unavailable, contradicted, and not applicable remain distinct states.
- Corrections: corrections@aiinsales.org.