FP&A
What Is a Driver-Based Forecast?
A driver-based forecast links every P&L line to an underlying operational driver such as units, AOV, conversion rate, headcount, or cost per order, instead of projecting each line as a percentage of revenue. This lets a CFO change one assumption, like raising Meta CAC from 48 to 58 dollars, and watch the model recompute paid customers, revenue, contribution, and EBITDA in 30 seconds without rebuilding it. The common failure is mixing models, leaving some lines driver-based and others percent-of-revenue, which produces nonsense scenarios.
A driver-based forecast links every P&L line to an underlying operational driver: units, Average Order Value (AOV), conversion rate, headcount, cost per order, instead of projecting each line as a percentage of revenue. This allows real sensitivity testing and scenario planning without rebuilding the model every time you want to test an assumption.
How it works
Instead of "marketing equals 30 percent of revenue," the model says "marketing equals $X paid spend across Y channels at Z Customer Acquisition Cost (CAC), producing W customers, who buy V times at U AOV." The output is revenue, but the inputs are the operational levers the team actually controls.
Common ecommerce drivers
- Revenue: units times AOV (or orders times AOV)
- Orders: paid customers plus repeat customers plus organic customers
- Paid customers: paid spend divided by paid CAC
- Repeat customers: prior cohort times retention rate
- Cost of Goods Sold (COGS): units times landed cost
- Fulfillment: orders times cost per order
- Payroll: full-time equivalent (FTE) count times fully-loaded cost per FTE
A worked example
A CFO wants to know what happens if Meta CAC rises 20 percent. In a percent-growth model, they'd have to reverse-engineer revenue and rebuild the whole forecast. In a driver-based model, they change paid CAC from $48 to $58 and the model recomputes: paid customers drop, revenue drops, contribution drops, EBITDA drops, all in 30 seconds. That speed is the unlock.
The most common mistake
Mixing models. Some lines driver-based, some percent-of-revenue. When you change an assumption, the driver-based lines respond but the percent-of-revenue lines don't, and the model produces nonsense scenarios. Either commit fully to driver-based or stay top-down. Don't half-ass it.
Frequently Asked Questions
what's a driver-based forecast in plain english?
Each P&L line is derived from operational drivers (units, AOV, CAC) instead of percent-of-revenue assumptions.
why is driver-based better than percent-growth forecasting?
You can test "what if CAC rises 20 percent?" in 30 seconds. Percent-growth models can't do that without a full rebuild.
what's the trade-off with driver-based forecasting?
Setup time and data discipline. Worth the investment above $5M ARR. Below $3M, top-down may be simpler.
Related Terms
- What is a 13-week cash flow forecast?
- What is scenario planning?
- What is sensitivity analysis?
- What is a rolling forecast?
Browse the full ecommerce finance glossary for every metric and money term a DTC operator needs.
Need a CFO to build a driver-based model? Talk to a CFO.