Forecast vs. projection: Which number should a CFO trust?

TL;DR: A forecast and a projection answer different questions, and treating them as interchangeable can cost a CFO credibility with the board or resources that never come back. This guide shows which number belongs in each decision and how Outreach keeps both numbers reconcilable, rather than letting them drift apart.

A board member asks whether the number in this quarter's board report is a forecast or a projection, and the presenter pauses because nobody labeled which was which.

Forecasts and projections answer different questions: a forecast is the current best estimate of what will happen, and a projection is what could happen if a specific assumption changes.

Confusing the two carries a real cost: a board mistaking an upside scenario for the forecast, or a resourcing plan built on a number that was never meant to be the base case. For CFOs deciding which number to present to the board, that distinction matters before the next quarterly review, not after.

Forecast vs. projection at a glance

To put things simply, trust the forecast when the organization needs to be accountable to a number, and turn to a projection when a decision depends on a choice.

A forecast answers what you now expect to happen, given current pipeline, historical performance, and no change of course, while a projection answers what could happen if one assumption changed: a win rate, a hiring plan, a price.

Neither number is inherently more accurate than the other; they simply serve different decisions, and the real risk is presenting a projection as though it were the forecast. That distinction holds up across five dimensions that determine how each number gets used.

The horizon row above is a common pattern, not a hard rule. Forecasts tend to be near-term and roll forward each period, while projections often stretch further out because they test a bigger question; a projection can model next month as easily as next year.

Whether a team builds a forecast top-down or through bottom-up forecasting, the assumptions behind each number, the expected conditions for a forecast, and a deliberately changed condition for a projection are what separate them, regardless of how far out either one looks.

What makes a number a forecast?

Before you can tell the two numbers apart in practice, it helps to nail down what each one is, starting with the forecast.

A forecast is management's current best estimate

A forecast is a forward-looking estimate built from historical performance, current pipeline data, and conditions already in motion, using whatever forecasting methods a team employs to turn that data into a number.

Its job is to be relied on as written, not the upside case, not the annual operating-plan goal, just the figure the business currently expects to hit. Whichever of the many forecasting software options a team runs those methods through, the number it produces still needs to meet that same bar.

Consider a team that closed 20 million dollars last quarter, has 24 million in qualified pipeline this quarter, and has seen nothing material change in the market or the sales process. The forecast for the current quarter should reflect that continuity, not a stretch target the team is chasing and not a number discounted out of caution.

What makes a number a projection?

A projection starts with the same underlying data and then asks a different kind of question.

A projection models a defined what-if

A projection is a forward-looking view built by deliberately changing one or more assumptions: a different win rate, a bigger sales team, a price increase, a new market, to answer what would happen under that specific condition. A good projection can be rigorous and useful, but it stops being trustworthy the moment someone presents it as the expected outcome rather than the scenario it represents.

Common CFO projection scenarios

Two examples show how this plays out inside a revenue organization:

• Headcount case: What happens to bookings if the company adds enterprise reps in the third quarter rather than the first?
• Pricing case: What happens to average contract value and retention if list price moves, and whether renewal rates hold once existing customers see the new number?

Building either scenario properly means adjusting win rate, pipeline coverage, and activity assumptions in a structured way, which is exactly what scenario planning covers in more depth.

The three-number problem: forecast, plan, and projection

The operating plan or budget is a third number that gets tangled up with forecast and projection just as often.

When the plan gets presented as the forecast

A budget or operating plan reflects what the organization committed resources to achieve for the period, not what is currently evidence-based and expected to happen. When a plan gets presented as though it were the forecast, the organization quietly rewards optimism over accuracy, and the forecast loses credibility every time the gap between plan and actual results widens.

Consider a business that set a 40 million dollar annual plan at the start of the year. If pipeline and conversion data support only 34 million, presenting the original 40 million figure as the current forecast hides a gap the board needs to see, not one worth smoothing over.

When the projection gets treated as the forecast

A projection built on a hypothetical, a new market or a price change sometimes gets carried into a board update or an operating review as if it were the base case. That is how an organization ends up committing real headcount, spend, and hiring timelines to a number that was never meant to represent the expected outcome.

A projection built around opening a new region, for example, assumes a ramp curve, a hiring plan, and a win rate the team has not yet proven, and carrying that projection into the operating forecast before any of those assumptions hold treats a hypothesis as a commitment.

Which number belongs in which CFO decision?

Once the three numbers are distinct, the practical question becomes which one belongs in front of which decision, a call that revenue operations teams and finance typically work out together.

Use the forecast for operating and accountability decisions

Lean on the forecast for the current-quarter revenue and bookings outlook, cash and liquidity monitoring, hiring and spend controls tied to the plan already in motion, and board reporting on expected performance. These are decisions that need a number updated as new evidence arrives, not a target or an upside case.

Use projections for strategic and contingent decisions

Use projections for strategic and contingent decisions. These decisions are inherently conditional, so the number backing them should be conditional too.

Use both for irreversible bets

For a large hiring plan or a market entry, start with the current forecast to understand the base case, then build a projection to model the decision under best-, expected-, and downside-case assumptions.

That sequence keeps an organization from approving irreversible spend based on an aspirational projection rather than the operating forecast that underlies it.

Consider a plan to double the enterprise sales team before the next fiscal year. The forecast shows what the current team can close without the expansion, and a projection built on ramp time, hiring cost, and expected win rate shows what the expanded team could add, with both numbers informing a single decision rather than one being mistaken for the other.

How CFOs should test whether a forecast is trustworthy

A forecast earns trust the same way any number does, by holding up when someone interrogates it, and there are a few tactics worth layering on top of boosting forecast accuracy more broadly.

One of the biggest takeaways from our annual revenue conference, Unleash, was that the goal isn't simply to produce a more accurate forecast—it's to create a forecasting process leaders can actually trust. When revenue teams have real-time visibility into pipeline health, and AI surfaces emerging risks early, finance and RevOps can spend less time debating the number and more time deciding what to do about it.

Ask what changed since the last forecast, and why

A regular pipeline review should ask what has changed since last time and whether that change can be traced back to something real in the pipeline or in the business. A forecast that jumped from 15 million to 18 million because three deals genuinely moved to late stage is defensible. The same jump with no corresponding pipeline movement is not, and that mismatch warrants investigation before anyone relies on the new figure.

Separate the evidence-based inputs from the judgment calls

Pipeline data, historical conversion rates, and current run-rate are evidence. A rep's confidence that a specific deal will close on time is judgment. Both matter, but a forecast worth trusting labels which inputs are which, rather than blending them into a single number nobody can unpack. This is also why CRM adoption across the team directly affects forecast quality: inconsistent inputs blur the line between evidence and judgment before anyone can separate them.

Trace the number back to the pipeline data behind it

Look underneath the roll-up total at deal-stage consistency, close-date movement, coverage ratio, forecast category, and renewal timing, the kind of pipeline inspection that determines whether the forecast on the slide reflects the pipeline in the CRM or a number that drifted away from it weeks ago. A forecast you can't trace back to specific deals cannot be defended in the next review.

How Outreach keeps the forecast and the projection talking to each other

The same discipline that keeps a forecast trustworthy also determines whether it stays connected to the projections built alongside it. Outreach, the only agentic AI platform for revenue teams, keeps both numbers grounded in the same pipeline data, so the forecast on the board slide and the scenario in the strategy deck can be reconciled instead of quietly drifting apart over a quarter.

Keeping every team's forecast on one governed number

Outreach’s Forecast Rollup aggregates forecasts hierarchically across reps, teams, and territories into a single number that finance can stand behind, and Parallel Forecasting keeps commit, best-case, and pipeline views distinct rather than blending them into a single blurred figure.

Building projections from the same pipeline data behind the forecast

Scenario Planner and AI Projection model revenue scenarios from the same underlying pipeline signals the forecast already draws on, so a projection stays reconcilable with the base number instead of living in a separate spreadsheet that drifts from it over time.

These capabilities strengthen the revenue and pipeline side of this equation. They do not replace cash forecasting or a full FP&A system, and they are not meant to. AI Projection surfaces a recommended what-if output for finance and RevOps to review; it does not replace the judgment call on which scenario to trust.

Omniplex Learning achieved 5 percent forecast accuracy, down from being off by 10 to 20 percent before it adopted these forecasting tools, a swing large enough to change how confidently its board treated the number.

Ask which number before you commit to it

The habit that matters most is not memorizing the difference between a forecast and a projection; it is asking the question before presenting either one. Before the next board meeting or resourcing debate, ask which number is in front of the room and what assumption sits underneath it.

A forecast says what to expect; a projection says what is possible under different conditions, and naming which one is on the table before committing real decisions to it is what keeps the numbers defensible in front of the board, the discipline Outreach's forecasting tools are built to reinforce.

‍

Frequently asked questions about forecast vs. projection

Is a projection more optimistic than a forecast?

Not inherently. A projection reflects whichever assumption changed, which can model a downside scenario just as easily as an upside one. The real risk is not that projections skew optimistic; it is presenting any projection as though it were the expected outcome rather than the conditional case it represents.

Should the board ever see a projection instead of a forecast?

Only alongside the base forecast, and only when it is clearly labeled as a scenario rather than the expected number. Presenting a projection in place of the forecast, without that label, is exactly the kind of confusion that puts credibility with the board at risk when the number does not hold.

Who owns the difference between forecast and projection, finance or RevOps?

Finance typically owns the forecast number itself, while RevOps owns the pipeline data feeding it, deal stages, close dates, and conversion history, a split covered in more depth in RevOps versus sales ops. Both functions share accountability for the forecast, since it is only as trustworthy as the data underneath it.

Can AI generate both a forecast and a projection from the same data?

Yes. AI Projection calculates a forecast from live deal signals, and the same underlying pipeline data can feed a projection built around a different assumption, win rate, hiring plan, or price change. A person still decides which scenario to trust and which number goes in front of the board.

‍