Revenue Metrics Cheat Sheet and Diagnostic Workbook

1. Write a metric contract before building the dashboard

A useful metric is a decision rule supported by a reproducible calculation. Its contract specifies the business question, unit, numerator, denominator, observation period, source system, owner and known limitation. Without that contract, two correct-looking reports can disagree because one counts people and the other counts opportunities. Begin with the decision, not the visualisation.

Use the original COUNT test: Cohort, Object, Unit, Numerator, Time. Ask which acquisition cohort is included, what object is counted, what unit the result uses, which records enter the numerator and when outcomes are observed. A metric that fails any part of COUNT should not determine spending. This is a practical review framework, not an accounting standard or a substitute for statistical analysis.

Create contracts for no more than a few decision-critical metrics first. Agree an “unknown” treatment rather than silently omitting incomplete records. Keep a version date when definitions change. If an earlier report used a different definition, show the break instead of presenting a smooth trend. COUNT improves clarity, but its limits remain important: a perfectly defined metric can still measure the wrong business objective.

2. Separate people, qualification, opportunities and customers

Lead volume measures records, not commercial value. Define a valid lead as a distinct usable contact with a relevant recorded action. Define qualification through explicit fit and business evidence. Define an opportunity through sales acceptance of a real buying process and a next step. Define a customer and its commercial value through an agreed contract or finance event. These definitions need local adaptation.

If several contacts contribute to one opportunity, a person-to-opportunity rate is a descriptive ratio rather than a literal probability for every individual. State that limitation. If one account creates two distinct opportunities, document whether both are legitimate separate buying processes. Do not delete complexity by merging unrelated deals solely to simplify a report.

The official CRM lifecycle guidance supports categorising contacts and companies, but the team's acceptance rules still require human agreement. Inspect a few records at every boundary before trusting a summary. A stage change without a reason or timestamp is hard to audit. Maintain explicit rejection and unresolved categories so incomplete follow-up cannot masquerade as poor fit. Keep opportunity amounts off the lead-volume line.

3. Calculate acquisition cost with the right scope

Cost per lead = included acquisition cost / valid leads. Cost per qualified lead = the same stated cost scope / qualified leads. Cost per opportunity = included cost / accepted opportunities. Customer acquisition cost, or CAC, divides acquisition cost by newly acquired customers. Every formula requires an explicit period and cohort treatment. A platform's media CPL and a fully loaded CAC are different measures.

Finance should approve allocations for shared staff, tools and agency work. Do not include all sales cost in one period and compare it with a channel's media-only estimate in another. If the organisation has a long sales cycle, report a cash-period view and a cohort view separately. Each answers a different question: what was spent now, and what did an acquisition group eventually cost?

For a zero-customer cohort, CAC is not zero. It is undefined or not yet observable, while expenditure remains real. Show the spend and unresolved outcomes explicitly. A very low CAC from one customer is unstable; disclose the count and commercial concentration. Cost metrics can identify affordability problems, but cannot alone prove that the activity caused incremental customers. Keep that inference separate.

4. Use payback and customer value without false precision

A simple gross-margin payback estimate is acquisition cost per customer / monthly gross profit per customer. Monthly gross profit is monthly recurring revenue multiplied by the agreed gross-margin rate, subject to finance's cost treatment. Payback in months is a planning measure; it is not the same as cash collection or accounting profitability.

Assume, fictionally, €8,000 acquisition cost, €1,000 monthly recurring revenue and an 80% gross margin. Monthly gross profit is €800; simple payback is ten months. If onboarding cost is outside the €8,000, include or disclose it. If billing is annual, the timing of cash receipt still differs from the monthly economic model.

These numbers are explicit fictional assumptions, not industry targets. Expansion and churn change the realised result. A simple lifetime-value formula using average revenue, margin and a stable churn rate can be misleading when churn varies by customer age or segment. Prefer observed retention cohorts and transparent scenarios when data is limited. Never hide uncertain lifetime value behind a precise decimal. A finance-approved downside case is more useful for a bounded spending decision than an impressive but fragile lifetime-value ratio.

5. Report pipeline scope separately from revenue

Pipeline is the commercial scope of open or accepted buying processes under a chosen stage definition. Define whether amount means first-year recurring contract value, total contract value, annual recurring revenue or another unit. Mixing those units makes totals and coverage ratios meaningless. Report the count and age of opportunities alongside the amount.

A weighted pipeline value is a model, not money that will certainly arrive. Stage probabilities should be calibrated with comparable historical outcomes where possible. If history is insufficient, label probabilities as assumptions and show an unweighted view. An opportunity that repeatedly shifts its close date should not retain confidence merely because its nominal stage is unchanged.

Ask sales to confirm a buyer-owned next step for material opportunities. Review stale deals and changes in amount before claiming pipeline growth. More pipeline can mean larger estimates, looser qualification or duplicate opportunities rather than better demand. Finance determines when a contract contributes to recognised revenue; marketing should not replace that rule with a platform conversion value. Keep contractual ARR, invoiced amounts, cash and recognised revenue distinct.

6. Reconcile recurring revenue, retention and expansion

For a defined recurring-revenue base, ending ARR = starting ARR + new ARR + expansion ARR − contraction ARR − churned ARR. Keep currency treatment, one-time services and contract timing consistent. ARR is an annualised recurring run-rate under the organisation's policy, not automatically recognised annual revenue. Finance should sign off the definition.

Gross revenue retention = (starting recurring revenue − contraction − churn) / starting recurring revenue. Net revenue retention adds expansion to the numerator. Exclude new customers from both retention numerators. Match the same starting customer cohort and period. If starting revenue is zero, the ratio is undefined.

Under these fictional assumptions, gross retention is 90%, net retention is 100% and ending ARR is €1,180,000. None of these figures is a benchmark. High acquisition can conceal weak retention in a top-line growth chart, so show the bridge. Segment results if customer groups have materially different contracts or onboarding experience, but avoid tiny cells that invite overinterpretation. When expansion belongs to another team's work, do not automatically attribute it to an acquisition campaign that happened years earlier.

7. Compare cohorts that have had time to mature

A cohort groups records by a common starting event, such as first valid inquiry in a calendar month. Later qualification, opportunities and customers are observed for that group. This prevents the common error of dividing this month's wins by this month's unrelated leads. The start rule must be stable and its handling of recycled contacts explicit.

Create a simple cohort matrix with acquisition month as rows and age since acquisition as columns. Populate cumulative accepted opportunities or customers for each age. Leave future cells blank. Do not replace an unobservable future cell with zero or extrapolate a final conversion rate without labelling the assumption.

Compare cohorts at the same age and inspect changes in segment, offer, sales capacity and definition. Different acquisition mixes can explain rate changes. Report the number of cases behind every ratio. Small counts create wide uncertainty, even when the decimal looks precise. A cohort view still cannot alone establish incremental channel impact; it improves temporal comparison, not causal identification. Keep open outcomes and their age visible in a separate queue, with an owner for the next action.

8. Keep attribution and causality separate

Attribution assigns observed credit according to a rule, such as first captured source, last captured interaction or a platform model. Causality asks what would have happened without an activity. A CRM source field and a platform report can both be internally consistent while answering different questions. Neither automatically identifies the incremental effect of advertising.

GA4 campaign URL parameters help identify referring campaigns. Google's enhanced conversions for leads and LinkedIn conversion tracking describe mechanisms for connecting later actions to advertising. These are measurement capabilities with implementation and privacy requirements. They do not remove every missing interaction or turn an association into a controlled experiment.

Do not sum attributed revenue from several platforms. They may credit the same customer under different observation rules and windows. Maintain a unique CRM opportunity and a finance-approved revenue record, then show each attribution view as a lens. Record model, window, timezone, conversion date convention and included events next to the report.

Use buyer-reported discovery and qualitative interviews as additional evidence, acknowledging recall limitations. Where the decision justifies the effort, consider a suitable holdout or geographic comparison with specialist support and an explicit analysis plan. Contamination, small samples and market changes can weaken the result. The copyable question is: “Which observed association are we reporting, and what additional evidence would be needed to call it incremental?” If the answer is unclear, use cautious language in the decision memo.

9. Run a fictional end-to-end diagnostic

FICTIONAL WORKED EXAMPLE — “Nordlicht Workflow” is an invented B2B SaaS business. Assume €24,000 programme acquisition cost for one cohort, 200 valid leads, 50 qualified leads, 20 accepted opportunities and five later won customers. Assume €12,000 first-year recurring value per won customer, 80% gross margin and no expansion or churn during the illustrated payback period. All inputs are assumptions, not client results, benchmarks or a forecast.

The resulting programme CPL is €120; cost per qualified lead is €480; cost per opportunity is €1,200; and programme CAC is €4,800. Qualification is 25%. If all 20 opportunities have a decision, five wins imply a 25% win rate. If ten remain open, the decided-opportunity win rate instead uses only decided records; do not quietly use the full 20 denominator and call the result final.

Now assume sales cost of €20,000 was omitted. Fully loaded CAC becomes (€24,000 + €20,000) / 5 = €8,800, and simple payback becomes eleven months. The channel did not suddenly deteriorate; the cost definition changed. Record the correction and revisit affordability using finance's agreed ceiling. If customer outcomes are still immature, report provisional values and continue observation rather than claiming the final ratios.

10. Use the monthly workbook and assign corrective action

Complete this worksheet with marketing, sales operations and finance. Begin by reconciling counts, then diagnose commercial transitions, then decide on investment. Do not spend the meeting debating a graph before establishing which records and time periods it represents. Keep a link to the approved source rather than creating uncontrolled personal-data exports.

Copyable review sheet

• Decision under review: ___; accountable decision owner: ___; review date: ___.
• Metric contract version: ___; acquisition cohort: ___; observed through: ___.
• Valid leads: ___; qualified contacts: ___; accepted opportunities: ___; open outcomes: ___.
• Won customers: ___; contracted ARR: ___; finance revenue definition and amount: ___.
• Included acquisition cost: ___; excluded costs: ___; allocation method: ___.
• Largest uncertainty: ___; downside scenario: ___; next evidence date: ___.
• Continue / modify / pause / stop: ___; specific action: ___; owner: ___; deadline: ___.

In the first 30 days, agree the dictionary and reconcile a small record sample. By day 60, publish a cohort view with visible age and unknowns. By day 90, review whether the definitions changed decisions and remove metrics that create work without improving judgement. This rollout is a practical sequence, not a guarantee of sufficient sales-cycle observation.

Pause spending expansion if submissions are missing, commercial stages cannot be reconciled or sales cannot follow up accepted inquiries. Change the offer or audience when buyer evidence identifies a repairable mismatch. Stop a test when its central assumption is contradicted and no affordable correction is justified. Hand data repairs to operations, qualification disagreements to the sales lead and cost or revenue disputes to finance. Every corrective action needs a verifiable completion condition, not merely a request to “improve reporting”.

Primary sources, checked 24 October 2026

• Use HubSpot contact and company lifecycle stages
• Understand HubSpot lead scoring
• Campaign URL parameters in GA4
• GA4 Measurement Protocol events reference
• About qualified leads and converted leads
• About enhanced conversions for leads
• Get started with LinkedIn Conversion Tracking

The GA4 sources support campaign URL parameters and event structures; Google Ads documents lead categories and enhanced offline matching. LinkedIn documents conversion data sources. HubSpot describes lifecycle categories and scoring. These capabilities replace neither your metric definitions nor evidence of incremental effect.