Most services companies are very good at forecasting revenue. They track their pipeline, apply weighted probabilities to opportunities, and produce quarterly revenue projections that are reasonably accurate. Then they miss their margin targets by 5 to 10 percentage points and wonder what went wrong. The answer is almost always the same: they forecast revenue but did not forecast costs with the same rigour. Revenue is only half the equation. Margin is the number that actually determines whether a project — and by extension, the company — is profitable.
Why Revenue Forecasting Alone Is Misleading
Revenue forecasting feels productive. It is concrete: you have a pipeline of opportunities, each with a value and a probability. You multiply, sum, and get a number. Leadership reviews it quarterly. Sales teams are compensated on it. Board decks feature it prominently. But revenue tells you nothing about project profitability.
A consulting firm in Brussels that wins a 500,000 EUR engagement at 20% margin and a firm that wins it at 40% margin both report 500,000 EUR in revenue. One makes 100,000 EUR. The other makes 200,000 EUR. Revenue is the same; the outcome is radically different. And because services companies operate with variable cost structures — unlike SaaS, where gross margins are relatively stable around 75-85% — project margins can swing from -5% to 55% within the same portfolio. The average might be 30%, but that average masks enormous variance. It is the variance that kills you, and revenue forecasting does nothing to expose it.
This is why margin forecasting has become the more critical discipline for services firms across Belgium and the broader European market. When your primary cost driver is labour and every project has a unique cost profile, forecasting revenue without forecasting cost is like navigating with half a map.
How Probabilistic Forecasting Works
Traditional margin forecasting uses point estimates: the project will cost X, generate Y in revenue, and therefore produce Z in margin. The problem is that point estimates are almost always wrong. They represent a single scenario in a distribution of possible outcomes. And in services delivery, uncertainty is not the exception — it is the norm.
Probabilistic forecasting replaces the single number with a range. Instead of saying "this project will generate 32% margin," it says "there is a 20% chance margin will be below 22%, a 50% chance it will be below 32%, and an 80% chance it will be below 40%." These are called p20, p50, and p80 estimates.
- p20 (pessimistic scenario). There is only a 20% chance the actual outcome will be this bad or worse. This is your downside risk — the number your CFO should be planning around.
- p50 (median scenario). The actual outcome is equally likely to be above or below this number. This is your best single estimate and should anchor your project planning.
- p80 (optimistic scenario). There is an 80% chance the actual outcome will be this good or worse, meaning a 20% chance it will be even better. This represents your upside potential.
The spread between p20 and p80 tells you something crucial: how uncertain the forecast is. A project with a p20 of 15% and a p80 of 45% has very high uncertainty — it could go either way. A project with a p20 of 28% and a p80 of 36% is much more predictable. This uncertainty signal is itself valuable for portfolio management: it tells you which projects need the most attention and oversight.
A Concrete Example: Where the Numbers Diverge
Consider a mid-sized IT services firm in Belgium running a fixed-price integration project. The contract is worth 320,000 EUR, scoped at 2,000 hours with a blended bill rate of 160 EUR/hour. The planned cost rate is 100 EUR/hour, giving a planned margin of 37.5%.
Revenue forecasting says: 320,000 EUR, done. That number will not change unless the client signs a change order. But the cost side is where reality diverges from the plan.
At the halfway point, the project has consumed 1,100 hours instead of the planned 1,000. Scope drift has added roughly 10% of unplanned work. The actual blended cost rate is 108 EUR/hour because a senior architect was pulled in to resolve a technical issue, replacing the planned mid-level developer. Actual cost to date is 118,800 EUR against a planned 100,000 EUR — already 18.8% over budget.
A point estimate EAC might project final cost at 237,600 EUR, yielding a margin of 25.8%. But probabilistic forecasting provides the fuller picture: the p20 is 19% (if scope drift continues at the current pace), the p50 is 26% (if the team stabilises), and the p80 is 31% (if remaining work comes in under estimate). The project's margin has dropped from a planned 37.5% to a likely range of 19-31%. That is a potential margin loss of 20,800 to 59,200 EUR — the kind of variance that, across a portfolio of 40 projects, can mean the difference between a profitable year and a flat one.
Why Services Companies Need Margin Forecasting
Services companies face a structural challenge that product companies do not: every project has a different cost structure. Manufacturing costs are predictable; scaling SaaS infrastructure follows known curves. But a consulting engagement is a unique combination of people, scope, timeline, and client behaviour. The cost drivers — labour hours, resource mix, rework, scope changes — vary from project to project and are difficult to predict in advance.
This is precisely why margin forecasting matters more than revenue forecasting for these firms. Revenue is contractually defined. Costs are emergent. Forecasting the emergent side is harder, but it is where the actual business risk lives. A European IT services firm running 50 projects per year with an average contract value of 200,000 EUR and a 5-percentage-point margin forecasting error is leaving 500,000 EUR per year on the table — not in lost revenue, but in unrealised profit.
The firms that forecast margin well share a common trait: they treat margin as an operational metric, not a financial reporting metric. They track it continuously during delivery, not just at month-end. They use it to make staffing decisions, scope decisions, and client conversations. And they use probabilistic ranges rather than point estimates, because they understand that precision without accuracy is worse than useless — it creates false confidence.
How Promapp Approaches Margin Forecasting
Promapp's approach to margin forecasting is built on two principles. First, forecasts should update automatically as operational data changes. When a time entry is logged, the EAC recalculates. When a task estimate is revised, the margin range adjusts. There is no batch process, no monthly reconciliation, no spreadsheet. The forecast is always current because it is derived from the same data model that drives project delivery.
Second, forecasts should be probabilistic by default. Promapp generates p20, p50, and p80 margin estimates for every active project, using a combination of current project performance data and historical patterns from completed projects — what we call "Project DNA." The system learns from your organisation's own delivery history: how much scope drift typically occurs, how accurate initial estimates tend to be, how much rate variance you experience. These patterns feed into the probabilistic model, making forecasts more accurate over time.
The result is that project managers and delivery leaders always have a current, probabilistic view of project profitability — not a stale point estimate from a spreadsheet that was last updated three weeks ago. They can see not just where margin is today, but where it is heading and how confident they should be in that trajectory. For services companies operating in the competitive Belgian and European market, that level of margin visibility is increasingly the difference between growing profitably and simply growing.