Separate the jobs.
Connect the numbers.
The Titan scenario starts with approximately $9.6 million of ARR and three customer segments. The modelling challenge is to keep source data, planning decisions and reporting outputs connected without making every change harder than it needs to be.
The approach gives each layer a clear responsibility: one place for shared data, one for planning drivers and one for the results leadership reviews.
One hub. Two focused spokes.
Data has an owner
The hub holds shared structures and loaded actuals. The planning and reporting layers reference them rather than creating competing definitions.
Reporting holds no inputs
Editable assumptions belong in the hub or planning layer. Reporting is designed to present and explain the resulting numbers.
A result you can reconcile.
The base illustration reconciles beginning ARR to ending ARR through new business, expansion, contraction and churn. All values below are whole USD.
| ARR movement | Amount |
|---|---|
| Beginning ARR | $9,619,493 |
| New business | +$2,025,000 |
| Expansion | +$1,655,474 |
| Contraction | −$291,087 |
| Churn | −$692,323 |
| Ending ARR | $12,316,557 |
| Bridge difference | $0 |
A zero bridge difference proves this arithmetic ties. In a production model, separate controls are also needed for source completeness, mapping, timing and business assumptions.
Try the scenario explorerChange a driver.
See the bridge move.
Adjust the two Forecast assumptions below while Budget stays fixed beside them. The bridge redraws against Budget, so the cause of every movement is visible before sign-off.
Adjust the Forecast. Budget stays fixed. Fictional sensitivity model; assumptions below.
| Driver | Budget | Forecast |
|---|---|---|
| Quota attainment | ||
| All segments | 75.0% | 70.0% |
| Monthly churn | ||
| SMB | 1.50% | 1.80% |
| Mid-Market | 0.80% | 0.80% |
| Enterprise | 0.40% | 0.40% |
| Monthly expansion | ||
| SMB | 0.60% | 0.60% |
| Mid-Market | 1.00% | 1.00% |
| Enterprise | 1.40% | 1.40% |
Scroll chart to explore →
Dashed outline: the other version. Figures in USD thousands, rounded so each column adds.
| Movement | Budget | Forecast | Difference |
|---|---|---|---|
| Beginning ARR | 9,619.5 | 9,619.5 | – |
| New business | 2,025.0 | 1,890.0 | −135.0 |
| Expansion | 1,655.5 | 1,647.0 | −8.5 |
| Contraction | −291.1 | −289.1 | +2.0 |
| Churn | −692.3 | −705.0 | −12.7 |
| Ending ARR | 12,316.6 | 12,162.4 | −154.2 |
The default Forecast ends $154.2K below Budget.
How this sensitivity model works
This browser illustration does not connect to Pigment. New business scales with quota attainment against a 75% Budget baseline. Expansion changes by $8.5K and contraction by $2.0K per $135K change in new business. Churn changes by $12.7K per 0.3 percentage-point change in SMB monthly churn. These are illustrative linear sensitivities, not cohort or monthly compounding calculations. Movements round to the nearest $100; ending ARR is their sum. Budget uses the displayed rounded values in USD thousands.
Protect the baseline.
Make change explainable.
Version design
Actual, Budget and Forecast each have explicit roles. Switchover timing belongs in the model design rather than being scattered across formulas.
Scenario consistency
Revenue assumptions should flow through the relevant statements and KPIs, with the same scenario definition throughout.
Review before sign-off
Compare driver changes and resulting outputs with the agreed baseline. Make unresolved differences visible before approval.
Traceable controls
Keep checks beside the outputs they validate. Distinguish arithmetic checks from source reconciliation and access testing.
What the interactive example does.
The homepage explorer changes three drivers in a single-period ARR bridge. It runs entirely in your browser with the fictional base values above.
- New business: an adjustable dollar amount, from $500,000 to $3,500,000.
- Expansion: a multiplier from 50% to 150% of the base expansion amount of $1,655,474.
- Churn: a multiplier from 50% to 150% of the base churn amount of $692,323.
- Held constant: beginning ARR of $9,619,493 and contraction of $291,087.
- Rounding: expansion and churn are rounded to the nearest dollar before the ending ARR is calculated.
The example does not model monthly compounding, customer cohorts, segment-level interactions or revenue recognition. It is designed to show how a small set of explicit assumptions can produce an explainable bridge.
What is needed before a Pigment implementation?
Start with the decisions the model must support, an accountable business owner, approved source samples, the planning calendar and acceptance criteria. Discovery confirms the dimensions, data quality, access and integrations before scope and milestones are agreed.
Bring your planning challenge.
We can discuss how this architecture maps to your data, users and reporting cycle—and what a sensible first phase would look like.
Discuss a similar buildTitan demonstration: FAQ
What is driver-based forecasting?
Driver-based forecasting builds the forecast from the operational inputs that cause the results, such as new customers, churn, price and headcount, instead of typing totals. When a driver changes, every dependent line updates.
How is ARR forecast in Pigment?
Titan forecasts ARR as a bridge: beginning ARR plus new business and expansion, minus contraction and churn, equals ending ARR. Each movement has its own drivers, and a control check confirms the bridge difference is zero.
Is Titan a real client model?
No. Titan is a fictional SaaS business used to demonstrate model architecture, versions and reconciliation. The browser explorer on this page is a simplified illustration.
Further reading: Rolling forecast: meaning, example and how to build one.