FP&A Forecasting & Strategies: 4 Stages You Need to Know

The short version

Most forecasts start strong and lose credibility by month three. The fix isn't more detail, it's a better structure that connects your P&L, balance sheet, and cash flow statement so you can test every assumption against actuals.

FP&A modeling is the practice of building financial models that connect your operating assumptions (headcount, revenue drivers, cost structure) to your income statement, balance sheet, and cash flow statement. Done well, it turns a static budget into a living system you can validate every month. Done poorly, it's a spreadsheet that looks polished but can't answer follow-up questions.

Josh Aharonoff, founder and CEO of Mighty Digits and creator of Your CFO Guy, joined Ramp for a 60-minute webinar on the FP&A modeling skills almost no one learns in school. You'll walk away with a four-level framework for forecasting, a method for projecting any balance sheet account, and the storytelling habits that get FP&A pros heard in the boardroom.

Most useful for: Finance and accounting professionals at companies between $1M and $500M in revenue who own (or want to own) the forecast.

What is FP&A forecasting?

FP&A forecasting is the practice of projecting a company's future financial performance (revenue, expenses, and cash flow) using historical results, operational drivers, and market signals. The point is to move from static annual budgeting to a continuously updated view of where the business is heading, so leadership can adjust strategy as conditions change rather than waiting for the next annual cycle.

A real FP&A forecast does three things at once. It connects the P&L, balance sheet, and cash flow statement into a single model. It traces every projected number back to an operational driver (headcount, conversion rate, units sold) rather than a target typed into a cell. And it lets you validate every assumption against actuals every month.

Without those connections, a forecast is a polished spreadsheet that can't survive its own follow-up questions. With them, it becomes a system you can defend to a board, refresh in days instead of weeks, and use as the basis for actual operating decisions.

Josh Aharonoff's four-level framework in the workshop further down shows exactly what each level of forecast maturity looks like, from a revenue-only projection at level one to a fully-actualized monthly BvA process at level four.

The main FP&A forecasting methods

FP&A forecasting isn't one technique. It's a set of methods you layer based on the question you're trying to answer.

1. Rolling forecasts

Instead of locking in a fixed annual budget, you maintain a forward-looking 12-month forecast that updates monthly. Each cycle, you fold in the latest actuals, refresh your assumptions, and extend the horizon by one period. The result is a continuous planning view that responds to actual performance instead of going stale six weeks into the fiscal year.

Josh's monthly BvA discipline in stage 5 that he mentions is rolling forecast practice in action, with a memorialized budget on one side and a live forecast on the other.

2. Driver-based forecasting

Don't forecast revenue directly. Forecast the operational drivers behind it: sales reps with ramp curves, conversion rates against ad spend, weighted pipeline, units sold at average order value. The output falls out of the math, and when actuals diverge, you can pinpoint exactly which driver broke. Josh's "inputs drive outputs" framing throughout the workshop is the practitioner's version of this method.

3. Scenario planning

Build the model once and stress-test it under multiple sets of assumptions: base case, downside case, upside case, plus variants for specific risks like a macro shock, a pricing change, or a customer concentration shift. The point isn't to predict which scenario hits. It's to know what your team's response is in advance, so the decision is faster when reality picks one.

4. Top-down and bottom-up

Top-down starts with a strategic target ("we need to hit $50M in revenue next year") and works backward through the drivers required to get there. Bottom-up starts with the operational reality (sales rep capacity, conversion rates, pipeline) and builds the number from the ground up. Strong forecasts use both. The top-down sets the ambition. The bottom-up checks whether the ambition is reachable.

5. Forecasting models by output

Revenue forecasts use historical trends and pipeline data. Expense forecasts split fixed and variable structures and build headcount as its own input. Cash flow forecasts ladder up from AR, AP, inventory, and working capital changes. All three have to tie back to each other, which is where Josh's BASE framework below ties the balance sheet together.

Why does the average FP&A forecast fall apart by month 3?

FP&A forecasts fall apart when they don't connect inputs to outputs across all three financial statements. If you're in FP&A, chances are nobody actually taught you that. You learned accounting, or finance theory, or how to use Excel. Then someone handed you a model and told you to keep it running.

Josh pushes against that inheritance. He treats FP&A forecasting as a system where the P&L, balance sheet, and cash flow statement are all connected. The model should turn inputs into decisions, not just display numbers.

That's why he keeps coming back to one idea: inputs drive outputs. If you're typing revenue targets into cells until you hit a top-down goal, you're building a target, not a forecast. A real forecast models the drivers behind revenue (sales reps with ramp periods, conversion rates against ad spend, weighted pipeline) and lets the output fall out of the math.

Once you connect inputs to outputs and historicals to projections, you can run budget versus actuals every month. You'll vet your assumptions against reality and walk into a board meeting with a story instead of a spreadsheet.

The 4 stages of FP&A forecasting you need to get right

Stage 1. A forecast with no actuals is barely a forecast

If you're forecasting revenue over the next 12 months with no costs, no balance sheet, and no actuals, you're at level one. That's where most teams start. A level four forecast (the goal) layers historical data underneath the projections so you can test every assumption monthly through budget versus actuals. Without actuals, you can't tell the board what happened, only what you guessed.

If you're at pre-seed or seed stage raising your first round, a revenue-only model may be enough, because investors are evaluating the business model. Past that stage, the gap shows up fast. The right way to build a level one model is to forecast the next 6–12 months in detail (the period leadership will hold you accountable to) and apply a simpler growth assumption beyond 12–18 months.

Josh learned this the hard way. After spending weeks building a complex multi-channel forecast for a client, he found himself 30 days later without access to the website analytics, click data, or conversion metrics needed to validate any of it. The detail was useless without the data feed to actualize it.

"I spent weeks and weeks, and I was so excited to share it with the client... 30 days later, a month later, I didn't have access to any of this information. I didn't understand how many people were visiting his website and how many people were clicking on this link. There was no way really to validate whether the assumptions I had in place were right or wrong."

Stage 2. Headcount breaks most level 2 forecasts

Headcount is probably the single largest cost on your P&L, and it's not one you can turn on and off. You might be forecasting it by taking the salary and dividing by 12.

The math fails the moment someone starts on the 15th, gets let go mid-month, or carries the roughly 20% burden in payroll taxes, health benefits, payroll admin fees, recruiting fees, and equipment that nobody added to the model.

If you're building a level two forecast, watch for three common failures:

• Salary divided by 12. Works for a contractor. Fails for an employee, who comes with a roughly 20% burden on top of base salary.
• No proration for start and end dates. If someone starts on the 15th, you can't book a full month of salary. The same applies to terminations mid-month.
• Treating commissioned roles like salaried ones. Salespeople with commission structures need a different build than salaried employees.

Headcount forecasting is complex enough to warrant its own input tab, prorated start dates, a full burden rate, and a separate build for each comp type.

"If someone starts on the 15th of the month or maybe even, like, the 30th of the month? You're now gonna be severely wrong with your projections if you're just going to take a twelfth. Instead, you have to prorate it, and the same is also true if someone has to be let go in the middle of the month."

Stage 3: Past $1M ARR, cash basis stops telling the truth

Accrual accounting separates the timing of cash from the timing of activity, and past $1M ARR, that distinction matters. It's the only way deferred revenue, AR, and inventory tell you what's actually happening in your business. The cash basis after that threshold hides too much.

Josh's rule of thumb is that if you're taking outside capital, you almost always need to be on accrual, and if you're bootstrapped, consider switching once you cross around $1M in ARR.

The mechanics of forecasting balance sheet accounts come down to the BASE framework:

• Beginning balance: Equal to last period's ending balance
• Additions: Events that increase the account
• Subtractions: Events that decrease it
• Ending balance: The result, which becomes next period's beginning

Deferred revenue makes the difference clear. Under cash basis, a $12,000 annual contract shows $12,000 in revenue when the cash arrives. Under accrual, you recognize one-twelfth of that amount each month, with the cash collected sitting in deferred revenue and drawing down as revenue is recognized.

Run the same BASE structure for AR, AP, inventory, accumulated depreciation, debt, and equity, and your balance sheet stays tied. If cash forecasting is a priority, the Cash Flow Playbook webinar walks you through building a cash flow model from scratch.

"The key reason for why you need to include a balance sheet is because when you have a P&L and balance sheet, you can now forecast not only your profitability, but more importantly, your cash flows."

Every balance sheet account follows BASE, and the 3 statements tie themselves.

Once BASE governs every balance sheet account, building the cash flow statement under the indirect method is a five-step exercise. In short, you're calculating how much cash moved by looking at how every non-cash account changed.

1. Duplicate the balance sheet tab. Wipe the formulas. Keep the structure
2. For every asset account, take last period's value minus this period's value. If an asset account grew, it means cash went out the door to acquire that asset
3. For every liability and equity account, reverse the sign. This period minus last period. If a liability or equity account grew, you paid with an IOU, took on debt, or received an investment instead of paying cash
4. Apply this to every account except cash. Because assets equal liabilities plus owner's equity, the net change of all non-cash accounts equals the change in cash. That's your plug
5. Link the cash plug back to the cash line on the balance sheet

Three P&L-to-balance-sheet connections work as one-to-one relationships. Net income flows into retained earnings. Depreciation and amortization expenses each accumulate into their corresponding balance sheet accounts. Other accounts influence each other (COGS and inventory, revenue and AR), but those aren't one-to-one and require deeper modeling.

The simplest error check, and the one that should end every modeling session, is whether assets equal liabilities plus owner's equity. If that ever stops being true, stop and fix it before you write another formula.

"To create a statement of cash flows, under the indirect method, all you're really doing is taking the net change in every balance sheet account to allow you to then get to your plug for cash... you could literally just duplicate your balance sheet, wipe all the formulas, keep the structure in place. Because assets equal liabilities plus owner's equity, the net change should equal as well."

Stage 4. The board needs explained variances, not zero variances

Budget versus actuals is the report that demonstrates your ownership of the finance function. The point isn't to nail every number. It's to walk into the board meeting and explain exactly why each variance happened and what you're doing about it.

You need a discipline you might be skipping. Once leadership signs off on the first version of the budget, save it down and never touch it again, because that version is now memorialized.

From that point forward, maintain a separate rolling forecast that you update monthly with new actuals and revised assumptions. Every month, compare the rolling forecast against the memorialized version. That comparison is your BvA report. Save down a new memorialized version before each board meeting and whenever the forecast deviates significantly from reality.

Josh once underestimated a client's cash flows by $3M over an eight-month horizon. Because he was running monthly BvA, he caught the error in the first month and presented a new plan before any damage was done. If you want to build the KPI dashboards and BvA reporting that make this sustainable, the Build 3 AI-Powered Dashboards in 60 Minutes webinar with Nicolas Boucher is a good next step.

"No, you don't need to have zero variances. Variances are expected. It's impossible to actually predict the future. But you do need to understand what happened and why did that differ from what I thought was going to happen…And when you do that, you can give confidence to the board of directors that you have everything under control."

Best practices for accurate FP&A forecasting

The methods above are the structural choices. The practices below are what separates a forecast that holds up at month three from one that doesn't.

1. Refresh on real-time data, not stale snapshots. AR, AP, headcount actuals, and spend data should flow into the model on a regular cadence (ideally weekly or daily, not monthly). When the source data lags, the forecast you presented in week one is already wrong by week three.
2. Collaborate across departments before each cycle. Sales and marketing know the demand drivers. Operations knows the supply constraints. RevOps knows the pipeline math. Without their input, your model captures last quarter's reality, not next quarter's. Schedule short working sessions with each function before you lock the forecast, not after.
3. Move beyond manual spreadsheets where the data is breaking you. Connected planning tools, data warehouses, and ETL pipelines reduce the manual reconciliation work that quietly destroys forecast accuracy. You don't need to rip and replace, but you do need to know which parts of your workflow are eating hours that should go to analysis.
4. Track the KPIs that actually drive the forecast. Identify the five to ten metrics that predict revenue and expense volatility for your business. Track them weekly. Build the forecast around them rather than around every line in the P&L.
5. Tie every assumption to an actual. No assumption should sit unvalidated. Run budget-versus-actuals every month against a memorialized budget, and document the variance explanations. Josh's stage 5 above walks through this discipline in detail, including the memorialized-vs-rolling pattern that makes it work.

How do you apply Josh's advice to get better at FP&A forecasting?

Each step below moves your forecast up one level on Josh's scale.

1. Audit where your current forecast sits on the four-level scale. Open the model. Is it revenue only (level 1), full P&L (level 2), balance sheet included (level 3), or historicals attached with monthly BvA (level 4)? Write down the answer and identify the single biggest gap.
2. Rebuild your headcount tab with prorated start dates and a full burden rate. Add roughly 20% on top of base salary for payroll taxes, benefits, payroll admin, recruiting fees, and equipment. Send the updated tab to your CEO and ask if the new total cost per employee (salary plus burden) changes the hiring plan.
3. Set up an error check tab anchored on assets equals liabilities plus owner's equity. If the equation breaks, the status flag surfaces on every tab. Add one more check for any place data passes from a source tab into a driver tab. This is the foundation that lets every later improvement compound.

Once those foundations are in place, your next step is stress-testing the model against different outcomes. The How to Build Scenarios Like a Wharton Program FP&A Leader webinar walks you through the scenario planning techniques that turn a solid forecast into a decision-making tool.

Pick the gap that costs you the most credibility today and fix that one first.

Final thoughts

"When working with accounting, you really need to invest in that relationship. Don't just be the one who every month says, where are my reports and why wasn't this done by this date? It's going to create a ton of friction which could ultimately sacrifice your ability to present the information that you need. So, invest in that relationship with accounting."

Josh's whole webinar is a quiet argument that FP&A is a relationship business dressed up as a spreadsheet business. The model gives you the numbers, but the relationships, with accounting, with department heads, with the CEO, and with the board, are what turn those numbers into decisions.

Pick the gap that costs you the most credibility today and fix that one first.

See how Ramp fits in

Your forecast is only as good as the data feeding it, and most of that data is locked in accounting systems, payroll platforms, and expense tools that close on different timelines.

You can close the books faster with Ramp, which automates expense categorization, vendor management, and accounting sync so your source tabs are accurate the day you refresh your forecast.

Common questions about FP&A forecasting and modeling

What is FP&A forecasting?

FP&A forecasting is the practice of projecting a company's future financial performance (revenue, expenses, and cash flow) using historical results, operational drivers, and market signals. The goal is a continuously updated view of where the business is heading, so leadership can adjust strategy as conditions change instead of waiting for the next annual budget cycle.

A real FP&A forecast connects the P&L, balance sheet, and cash flow statement, traces every projected number back to an operational driver, and validates assumptions against actuals every month.

What are the main types of FP&A forecasting?

The main methods are rolling forecasts, driver-based forecasting, scenario planning, and the combination of top-down and bottom-up approaches. Most strong forecasts use a layered version of all four, plus output-specific models for revenue, expense, and cash flow.

What's the difference between forecasting and budgeting in FP&A?

A budget sets the strategic target leadership commits to at the start of the year. A forecast is the rolling view of where the business is actually heading, updated as new actuals come in. Once a budget is approved, it should be memorialized and not edited. The forecast then runs alongside it, and the gap between the two becomes your budget-versus-actuals story.

Most teams collapse these two artifacts into one moving document and lose the ability to explain variance as a result.

What is FP&A modeling, and how is it different from regular budgeting?

FP&A modeling is the practice of building financial models that connect your operating assumptions (headcount, revenue drivers, cost structure) to your income statement, balance sheet, and cash flow statement. Where budgeting sets a target, FP&A modeling builds the mechanism that explains how you get there, tests whether you did, and updates the story when reality diverges.

The difference shows up most clearly in Josh's four-level framework, where a budget is roughly a level one or two artifact, while a model that earns board trust operates at level three or four, with actuals attached and variance explanations ready.

When should we switch from cash to accrual accounting?

Josh's rule of thumb: if you're taking outside capital, investors will almost always require accrual. If you're bootstrapped, consider switching once you cross around $1M in ARR. Below that threshold, cash basis is faster to close and you won't miss much insight. If you have deferred revenue, AR, or inventory of any size, you're already past the point where cash basis tells the full story.

What KPIs should you track to support FP&A forecasting?

The strongest KPIs are the operational drivers that actually move revenue and expense for your business, not generic finance ratios. Examples include weighted pipeline coverage, customer acquisition cost, ramped sales rep capacity, conversion rates by channel, gross margin by product line, headcount with prorated start dates, and working capital metrics like DSO and DPO.

The five to ten that matter most are the ones a 10% change in would visibly move your forecast.

What is the fastest way to prevent errors in a complex model?

Build an error check tab and route every check to it. The non-negotiable check is whether assets equal liabilities plus owner's equity. Add a check anywhere data moves from a source tab into a driver tab, comparing totals on both sides. Surface the status flag on every tab so a broken model announces itself the moment it breaks.

Josh compares it to a roll call on a school field trip, you count the kids before they get on the bus and count them again when they get off.

How do you turn financial model outputs into a story the board will act on?

Start with the variance that matters most, explain what caused it, and state what you're doing about it. Lead with the business impact, not the accounting mechanics. Frame your numbers around the questions your board is already asking (runway, growth rate, hiring capacity) and connect each data point to a decision they need to make. The How to Use AI to Turn Raw Data into Executive-Ready Insights webinar covers data storytelling techniques that help you present analysis in a way that drives decisions.