Let’s be honest. You didn’t get into finance to fix broken formulas at 2 a.m. You want to answer the big questions that actually move the needle for your business.
Driver based planning flips the script on traditional modeling. Instead of hard-coding every line item, you build your model around the operational factors that matter, such as customer count, churn, or average order value. You change one driver and your entire forecast updates automatically. You stop searching through tabs and start analyzing impact.
We’re here to help you master this approach. This guide breaks down exactly how to start driver based planning without getting bogged down in the weeds. You’ll explore real-world examples, learn to pick the right drivers, and leave with a practical workflow you can use right now.
Why driver based planning changes the game
Traditional forecasting often feels like running on a hamster wheel. You recreate financials from scratch every month. You update accounts manually, roll numbers forward, and lose hours to maintenance. It’s slow work. One assumption changes and you’re stuck chasing updates across dozens of isolated cells.
Driver based planning does it differently. You identify the operational levers that matter most, such as customers, pricing, churn, and headcount. You connect these with logic so a single change flows through your P&L, cash flow, and balance sheet immediately.
The benefits speak for themselves. FP&A teams often see forecast accuracy jump by 50% or more. You get to spend your energy on strategic analysis rather than data entry. Scenario planning becomes a breeze when you only need to adjust a few key drivers instead of hundreds of static cells.
Despite the perks, only 9% of organizations use fully driver based models. It’s not for lack of interest. Implementation feels daunting. Teams often hit walls with data quality or struggle to select the right metrics. We’ll help you clear those hurdles and build a model that works as hard as you do.
Building a clear financial picture with drivers
Driver based planning shows how cause and effect connect. Operational choices, like bringing in customers or shipping orders, influence key KPIs, which then shape your financial results.
Think of it as a pyramid:
- At the base, you have operational inputs controlled by teams every day.
- The middle has derived metrics that summarize the inputs.
- Your financial statements sit at the top, powered by the layers below.
For a SaaS company, customer count and ARPU drive MRR. MRR leads to ARR. That feeds into revenue. Churn reduces customer counts and impacts the whole chain. CAC and sales rep quotas shape how many new customers you bring in. Those roll right into your main drivers.
Make these links clear. Drivers in Runway are timeseries building blocks. Define, connect, and see how each piece moves your model. Change a driver and the update happens instantly. You stop manual updates. The model keeps you current.
Build a simple driver tree
Start small. Pick your top two or three drivers - the factors that make the biggest difference. For most, that means:
- Volume (customers, orders, transactions)
- Pricing (ARPU, AOV, take rate)
- Retention (churn, renewal rate)
Layer in detail as you go. If customer count is key, what drives it? New customers and churn. If new customers are a driver, consider marketing spend or sales capacity.
You don't have to model everything. Focus on the factors that explain most of the swings in your results. A solid SaaS model might look like:
- Customer counts by tier
- Churn and upgrade rates
- Acquisition channels
- Fixed and variable costs
- Headcount plans
This is enough to forecast revenue, gross margin, OpEx, and your runway.
Shape your tree so it's shallow and wide, not deep and tangled. Aim for a few main drivers, each with a handful of inputs. If you find yourself building endless layers, step back. You need simplicity.
What makes a good driver?
Not every metric belongs as a driver. Good drivers share four traits:
- Controllable: You can influence it through decisions, like pricing, hiring, or marketing spend. Not GDP growth. Use macro factors as scenarios, but focus your core on what you control.
- Measurable: You can track it over time. If you can't count it, you can't plan with it. "Customer count" works. "Brand awareness" doesn't, unless you've assigned a real metric.
- Leading: It acts as an early signal. Pipeline velocity predicts bookings. Bookings indicate future revenue.
- Stable relationship: It consistently ties to outcomes. If a metric swings without a pattern, it won't help you forecast.
When you choose drivers, apply these tests. Drop any that fail two or more. Stay focused on the strongest levers.
Choose drivers for your business model
Your best drivers depend on your business. SaaS teams care about churn and ARPU. Ecommerce tracks AOV and returns. Marketplaces look at GMV and take rate. Here’s how it breaks down by model:
SaaS or subscription-based
SaaS means recurring revenue is king. Focus on customer count, ARPU, churn, and bookings. A typical SaaS forecast starts with customers by tier, applies churn and upgrade rates, adds customers from self-serve and sales, multiplies by ARPU, and gets MRR.
Sales-led growth? Add in number of reps, quota per rep, ramp speed, and attainment. Don't bake in 100% attainment. Set realistic assumptions.
Net revenue retention (NRR) gives early warnings; whether customers stick and spend more or start to churn. Use NRR for cash forecasting that doesn’t rely on guesswork.
For costs, break out:
- Fixed: Salaries, office, tools
- Variable: Hosting per customer, payment fees
- Growth: Marketing, commissions
Tie headcount to hiring dates and commissions to bookings so costs update as your plan changes.
Usage-based or product-led growth (PLG)
Usage-based models link revenue to actual customer consumption. Main drivers are:
- Active users
- Usage per user
- Pricing per usage unit
If you use freemium pricing, add conversion rate (free to paid) and ARPPU (average revenue per paying user).
ARPPU is usually two to five times higher than ARPU in freemium models. Measure both. ARPU is about broad monetization, ARPPU is about value from paying customers.
Cohort ARPU is also valuable. It tracks how groups that started together grow over time—by one, three, six, or twelve months. If your revenue is bumpy, use an average to spot trends.
Ecommerce and direct to consumer (DTC)
Ecommerce needs drivers like:
- Traffic
- Conversion rate
- AOV
- Return rate
Revenue is orders times AOV, minus returns and discounts. If conversion rates drop, ad spend adjusts automatically in a connected model.
Key cost drivers:
- CAC (or ROAS)
- Fulfillment cost per order
- Inventory
COGS is beginning inventory plus purchases, minus ending inventory. If you're growing, watch days of supply so you avoid cash strain.
Stick to the 80/20 rule. Model your most important SKUs individually. Group the rest. You capture most of the picture without drowning in detail.
Marketplace
Gross merchandise value (GMV) and take rate drive everything. GMV is transaction value. Take rate is your slice. Together, they show if your business is thriving.
GMV is sale price times transaction count. Net GMV removes returns and chargebacks. Revenue is net GMV times take rate. Subtract variable costs—payment processing, support, and SMS—for your contribution margin.
Break out GMV:
- By location
- Buyer vs. seller fees
- Promotions
If you operate in multiple countries, pay attention to how fees and discounts change the bottom line. Buyer fees affect conversion. Seller fees impact retention.
Services organizations
Services focus on time and utilization. Revenue is billable hours times billing rate. Fully burdened labor rate is the true hourly cost. Salary plus benefits, taxes, and overhead.
If someone earns $25 per hour, the real cost may be $35-$38 after everything. That's a huge jump. Underestimate here and you'll miss by 30-40%. That small gap compounds over time.
For consulting, track:
- Utilization (billable vs. available hours)
- Project win rate
- Backlog
Keep utilization near 75%. Push higher and people burn out. Too low and you leave money unearned.
Avoid false precision in your forecasts
It's easy to think more detail equals more accuracy. But false precision creeps in when you report forecasts to the dollar, even though your data is much fuzzier. Overly precise numbers mislead your team.
Here's how you keep it real:
- Use ranges and show confidence levels. Make uncertainty clear. Don’t present a single number as the only possibility.
- Keep your driver sets small. Limit scenarios to three to five. Too many can overwhelm. The same goes for drivers. Focus, then expand as data justifies.
- Backtest and adjust often. Check how well your model matches reality. Revisit driver links. If a driver’s relationship has shifted, update it.
- Don't overcomplicate. Before adding a new variable, make sure it truly adds clarity and accuracy.
- Round numbers to match your input uncertainty. If drivers could swing 10%, don’t show decimals. Match your outputs to the level of confidence you have in your inputs.
Use a lightweight workflow to launch driver based planning
Here’s a clear way to get started:
- Pick your core drivers. Choose two or three with the highest impact for your business, customers, ARPU, and churn for SaaS; orders, AOV, and return rate for ecommerce. Define these and set them up with simple formulas.
- Connect data sources. Identify where each driver lives. CRM, accounting, or product analytics, and automate the feed where possible. Organize drivers into clear tables and consolidate data.
- Build your driver tree. Link main drivers to derived metrics and financial outcomes. For example:
Encode these so changes ripple automatically. - Revenue is customers times ARPU
- Customers is last month's customers plus new minus churned
- Churned customers is last month's count times churn rate
- Backtest. Compare your model to real results. If gaps are big, revisit your selections and formulas.
- Create scenarios. Scenarios are edits on top of your main plan. Shift a few drivers. Bump churn, lower CAC, or add sales reps. Compare new forecasts side by side.
- Set a regular cadence. Update drivers monthly or quarterly. Keep scenarios fresh by syncing in new data. Assign owners and document where numbers come from.
- Recalibrate often. Test driver links. Use an "Unused Drivers" view to spot drivers you aren't using and clean them up. Keep things lean.
Watch out for common driver based planning pitfalls
Teams hit predictable snags. Here’s what to watch for and how to steer clear:
- Disconnected models. If drivers live in one spreadsheet and actuals in another, you’re just doing more spreadsheets. Most models report the past, not simulate the future. Plug drivers into live data so updates happen automatically.
- Manual spreadsheet overload. Many teams spend over half their time patching formulas, not analyzing drivers. Native driver logic frees you up to ask “what if?” instead of fixing errors.
- Missing total labor cost. In services, focusing on base salary could mean you’re underestimating by 30-40%. Labor usually makes up 60-80% of total costs, so even a small offset adds up. Use fully burdened rates.
- Too many scenarios. Three to five is enough. Each should answer a real question like what if churn rises or hiring moves faster.
- Stale drivers. Drivers lose value over time. Your model should shift as the business changes. Keep scenarios current by updating monthly or quarterly. Test your driver links regularly.
- Poor data quality. Great inputs build great forecasts. Assign ownership to data. Document sources. Automate where you can. High quality data means sturdy, trusted models.
Action steps and next moves
Driver based planning replaces line items with levers. Change a driver and numbers update everywhere. It's quicker, clearer, and keeps you in control. You stop focusing on data entry and spend more time making decisions.
Start simple. Grab your top two or three drivers. Connect them to live data. Build your driver tree. Test with old results. Then build from there. Add scenarios. Bring in more drivers when you see the value.
Skip false precision. Use ranges instead of one number. Keep things tight. Test often. Match your rounding to what your inputs justify. A $200K overrun might point to bad planning or smart growth. Trace results back to drivers and you'll see the story.
If you're ready to stop firefighting and start leading, talk to us at Runway. See what collaborative, driver based planning feels like. Run scenarios. Compare results. Move faster and plan better.