A revenue forecast projects income over a future period using historical results, pipeline data and assumptions about growth, churn, seasonality and pricing. It is not a cash forecast: revenue is recognised when it is earned, which is often not the month the money arrives.
What a revenue forecast actually projects
A revenue forecast takes a baseline, usually the last twelve months of recognised revenue, and grows it forward by period, most commonly monthly for the coming year and quarterly beyond that. In a subscription business the baseline splits into existing recurring revenue, expansion, contraction and churn, with new business layered on from pipeline. In a transactional business it splits by volume and average price.
One distinction to get right early: forecasting revenue is not the same as forecasting cash. Revenue is recognised when it is earned under the AASB accounting standards, which is frequently not the month the money arrives. An annual contract billed upfront and a monthly one at the same price look identical on a revenue forecast and nothing alike on a cash forecast.
The data problem is the real problem
A revenue forecast pulls from the CRM for pipeline, the billing system for recurring revenue, and the ledger for what actually landed. In theory those three reconcile. In practice they use different definitions of the same word, and that is where the forecast quietly dies.
Senior finance leaders recognise the pattern immediately: inherit non-standardised reporting styles, processes and data definitions across different entities and systems, and you lose the ability to see the bigger picture at all.[1] If sales counts a deal as closed on verbal agreement and finance counts it on signature, the two revenue forecasts will differ by a quarter and both parties will be confident. Nobody is wrong. There is no agreed definition.
Fixing this is unglamorous and it is the highest-return work in the whole exercise. Agree the definitions, agree the source of truth per field, and only then argue about growth rates.
The assumptions are the forecast
Four assumptions carry almost all the variance in a revenue forecast: the growth rate applied to new business, the churn or retention rate applied to the existing base, seasonality, and price. Everything else is arithmetic. Which means the model should let you change exactly those four in one place and watch the outcome move, and it should record who changed them and why.
| The assumption | The question that should be answerable on the spot | |
|---|---|---|
| Growth rate | Applied to new business by period | Where does this rate come from, and what did we actually deliver against the same assumption last quarter? |
| Churn or retention | Applied to the existing revenue base | Is this measured on customers, on dollars, or on logos, and is it the same definition sales and the board are using? |
| Seasonality | Shapes revenue across the year | Do we have enough history for this to be a real pattern, or are we modelling one unusual year twice? |
| Price | Average deal size or unit price | Is a planned increase in the model, and has anyone tested it against renewal risk in the existing base? |
When finance ownership is genuinely in place, the change founders describe is that reporting gets clearer, forecasts become more believable, board prep gets less chaotic, cash conversations get more accurate, and hiring plans get properly tested before they become commitments.[2] Believable is the word that matters. A revenue forecast is a claim, and a claim needs someone standing behind it.
The most reliable way to make a revenue number defensible is to build it from pipeline upward, which is what bottom-up forecasting is for.
Who should be able to change them
Access control in a forecasting tool is a real feature, and it is usually solving an organisational problem rather than a security one. If the sales leader can edit the growth rate, the forecast becomes a target. If nobody can, it goes stale. The workable answer is that sales owns the pipeline inputs, finance owns the conversion and churn assumptions, and one named person owns the output.
That named person needs a commercial skill set, not just a finance one. The requirement I see written into CFO briefs is someone who can lift commercial acumen across the business and work effectively with sales, operations and product stakeholders.[3] A finance leader who cannot sit with a sales team and challenge a conversion rate without starting a fight will not produce a useful revenue forecast, whatever the model looks like.
On timing, in the businesses I work with the first CFO usually lands at around 50 heads and roughly $10 million ARR, typically Series A or Series B.[4] Before that, the forecast belongs to the first finance hire, and it should already be visibly improving. By about day 60 into that hire, founders report reports arriving proactively, an easily explainable cash position, realistic forecasts, and noticeably fewer finance questions coming back to them.[5]
If the revenue forecast is the thing the board keeps pushing back on, the real question is when to hire your first CFO.
Common questions
What is a revenue forecast?
A revenue forecast projects future income over a defined period using historical performance, the current sales pipeline and stated assumptions about growth, churn, seasonality and price. Subscription businesses typically separate existing recurring revenue, expansion, contraction and churn from new business won out of pipeline.
How is a revenue forecast different from a cash forecast?
Revenue is recognised when it is earned under the accounting standards, which is often not the month the cash arrives. An annual contract billed upfront and a monthly contract at the same annual price look identical on a revenue forecast and completely different on a cash forecast. Growing companies that track only one of the two get surprised by the other.
Why do revenue forecasts keep missing?
Usually not because the maths is wrong. The two recurring causes are inconsistent data definitions between the CRM, the billing system and the ledger, so the same word means different things in different systems, and assumptions that nobody is accountable for. Fix the definitions first, then argue about growth rates.
Who should own the revenue forecast?
Sales should own the pipeline inputs, finance should own the conversion and churn assumptions, and one named person should own the published output. That person needs commercial credibility as much as technical skill, because most of the job is challenging a conversion rate with a sales leader without it becoming an argument.
References
- What keeps coming up: inheriting non-standardised reporting styles, processes and data definitions across different entities, leading to an inability to see the bigger picture.
- Where I land on this: reporting gets clearer, forecasts become more believable, board prep gets less chaotic, cash conversations get more accurate, and hiring plans get properly tested before they become commitments.
- From the CFO briefs Tom Hunter writes: clients repeatedly ask for a strong combination of commercial skill set and personality, someone who can provide an overall uplift in commercial acumen across the business and work effectively with sales, operations and product stakeholders.
- What the numbers look like here, from running a specialist finance recruitment firm for Australian tech, fintech and deep tech companies, as discussed on the Honest Wealth Builders podcast: a business with about 50 heads and around $10 million ARR, typically at Series A or Series B, will hire its first CFO.
- A pattern I see repeatedly: by day 60 founders start to feel the impact, observing reports produced proactively, an easily explainable cash position, realistic forecasts and less last-minute board preparation scramble, with the most noticeable change being fewer finance questions directed at them.