Forecast the Whole Growth System
Task
Formalize forecasting and executive GTM dashboard.
Summary
Bring pipeline, recurring revenue, churn, expansion, acquisition cost, payback, and channel mix into one view.
A sales forecast is only one part of growth
A company can hit its bookings target while recurring revenue misses plan because customers start late, fail to activate, contract, churn, or require more support than expected. It can also miss a monthly sales target while the existing customer base expands enough to produce a healthy result.
Forecast the connected system: demand, pipeline, sales, onboarding, activation, recurring revenue, retention, expansion, cost, and capacity.
Start with controlled definitions
Every metric needs a written definition that states:
- the business question it answers;
- formula and population;
- event date and measurement period;
- inclusions and exclusions;
- currency and contract rules;
- source systems and owner;
- refresh cadence;
- segmentation;
- reconciliation method;
- known limitations and decision triggered.
Do not treat annual recurring revenue as recognized revenue or cash. Do not compare channel customer acquisition cost when channels assign labour, partner payouts, or influenced opportunities differently. Freeze definitions before comparing actuals with forecasts.
Connect the forecast layers
| Layer | Forecast question | Inputs |
|---|---|---|
| Demand | How much suitable buying activity will enter? | Channel activity, response, accepted-opportunity rate, seasonality |
| Pipeline | Which opportunities will progress and when? | Stage, value, close date, evidence, next step, historical movement |
| Sales | What will be won? | Opportunity forecast, conversion, cycle, price, discount, capacity |
| Onboarding | When will customers become active? | Start dates, implementation capacity, completion and exception rates |
| Recurring revenue | What recurring value will be active? | Opening base, new activation, expansion, contraction, churn, timing |
| Customer economics | What will growth cost and return? | Acquisition, partner, onboarding, hosting, support, and gross profit |
| Capacity | Can the company operate the result? | Sales, implementation, support, product, and partner constraints |
Maintain both a company total and channel views that reconcile to it. Preserve the original forecast, later revisions, management adjustments, and actual result so the company can learn which process steps improve accuracy.
Forecast recurring revenue as movements
Use a bridge rather than one unexplained total:
Forecast each movement from its own evidence. New recurring revenue depends on sales and activation timing. Expansion depends on eligible customers and qualified changes. Contraction and churn depend on renewal dates, use, value, risk, and commercial conditions.
Separate immature cohorts. A channel that recently produced many customers may appear successful before onboarding, support, retention, and renewal evidence has had time to emerge.
Measure error and bias
Forecast accuracy should be assessed by horizon, segment, channel, and type of movement. At minimum retain:
- forecast minus actual for direction and size;
- absolute error so positive and negative errors do not cancel;
- bias over several periods to expose persistent optimism or conservatism;
- a weighted percentage or scaled error when comparing differently sized groups;
- forecast value added to test whether overrides or extra process steps improve the result.
No one accuracy measure works in every setting. Percentage errors become unstable when actual values are zero or very small. Pair the number with an explanation of the miss and the decision it affected.
Run one joined review
The weekly review should ask:
- What changed since the previous forecast?
- Which evidence supports the change?
- Where do channel, sales, onboarding, customer, or finance views disagree?
- Which assumptions are repeatedly wrong?
- What action could still change the result?
- Who owns that action and when will it be reviewed?
Review exceptions and material movements, not every dashboard tile. Keep targets separate from forecasts: a target states the desired result; a forecast states the result currently supported by evidence.
Use the dashboard to make decisions
The executive view should show the opening recurring base, forecast movements, pipeline coverage and quality, acquisition cost, payback, activation, retention, expansion, support burden, channel mix, forecast error, known constraints, and assigned actions.
Every total should drill into a defined population and source. If a number cannot be explained, mark it as unresolved rather than presenting false precision.
What must be true before relying on the forecast
The forecast is dependable when metric definitions are controlled, channel and company totals reconcile, pipeline and recurring-revenue movements connect to customer events, uncertainty and cohort maturity are visible, forecast changes are preserved, and review actions have owners. The goal is not perfect prediction. It is earlier, better decisions about spending, capacity, customer risk, and where the growth system needs work.