Subscription Renewal Forecast Bridge

A subscription business knows which contracts are up for renewal but cannot know who will renew or change plans. A renewal forecast bridge connects a starting recurring-revenue base with a dated expected ending base.

Start with a clear measurement date, using active subscriptions and their recurring amount at that point, and do not mix a month-end snapshot with transactions added later without recording the difference. Define recurring revenue consistently, since monthly recurring revenue and annual recurring revenue are run-rate measures under chosen rules, not cash collected in a month, and one-time setup fees should not silently enter the recurring base.

Identify each renewal cohort by grouping subscriptions by contractual renewal date, product, customer segment and currency as needed, because an annual plan should not be treated as if it renews monthly. Separate known future changes from assumptions, as a customer who has already sent a valid cancellation notice differs from one whose account team thinks renewal is unlikely, and label the evidence behind each line.

Committed MRR forecasts may project scheduled upgrades, downgrades and cancellations, and ChartMogul describes this as a view of scheduled changes; an assumption-based renewal bridge is a different, more uncertain planning layer. Map the starting value: suppose $500,000 in annualised recurring revenue is associated with contracts renewing in the coming quarter, kept tied to a frozen cohort list.

Estimate retained value carefully, because if the illustrative cohort is expected to retain $420,000 the forecast churn or non-renewal is $80,000, which should not be claimed as a signed commitment. Add expansion and contraction separately, since a renewal may retain the customer but change quantity or price and showing the movements avoids hiding a downgrade inside a simple yes-or-no renewal count.

A simplified bridge might read: opening renewal cohort $500,000, less forecast loss $80,000, less contraction $20,000, plus expansion $30,000, equals $430,000 forecast post-renewal annualised value, which does not include new customers outside the cohort. Watch overlapping categories, because a full cancellation already removes the entire contract value and its expected seat reduction should not also be counted as contraction.

Distinguish gross and net retention, as gross retention excludes expansion in many commonly used definitions while net retention includes it, so write the numerator and denominator rather than relying on a label alone. Consider renewal timing, since a contract may renew late, lapse temporarily or switch to a different billing cycle, and decide how each state enters the bridge and explain timing differences.

Track the sales process, because renewal notice dates, procurement steps and customer approvals can move outcomes and a verbal intention should not be treated like a signed order. Use probability with care: multiplying each contract value by a renewal likelihood can help scenario planning, but those probabilities should be checked against past cohorts, not assigned only by optimism, and a base case, downside and upside can show sensitivity to a few large contracts, so state which contract decisions drive the difference.

Connect the bridge to billing records, since a contract signed at a new price may not yet appear in the subscription system, and do not treat the forecast as recognised revenue, because accounting recognition can differ from billings and recurring run rate, so finance needs a separate revenue schedule under the relevant policy. Measure forecast error after the period by comparing forecast retained, lost, expanded and contracted value with actual outcomes from the original cohort while preserving the earlier forecast, and investigate misses by cause, since a late procurement decision, product problem or incorrect contract date calls for different remedies; for owners, the bridge turns a renewal target into traceable movements and is most useful when confirmed contract changes and uncertain judgments remain visibly separate.