A sales forecast is a projection of the revenue a sales team expects to close within a defined time period — typically a month, quarter, or fiscal year. For SaaS companies, the forecast is one of the most important operating documents in the business. It drives hiring decisions, marketing spend, board reporting, and investor confidence. Getting it right matters enormously. Getting it wrong, repeatedly, erodes trust at every level of the organisation.
This guide explains what a sales forecast is, how SaaS teams build one, why they so often go wrong, and how AI is changing the accuracy and reliability of forecasting for modern revenue teams.
At its core, a sales forecast aggregates the expected close value of open opportunities within a time period. But the inputs that feed that number — and the process by which they're validated — vary significantly between teams.
The most common forecasting inputs in SaaS include: opportunity stage and weighted probability, rep-submitted commit and best-case categories, deal size and expected close date, historical win rates by stage and segment, and MEDDPICC qualification scores. Most CRMs apply a weighted pipeline model — multiplying each deal's value by its stage probability — to produce a forecast number. The problem is that stage probabilities are usually set once and never revisited, and rep-submitted commits are notoriously optimistic.
The result is a forecast that looks precise but is actually a collection of guesses dressed up in a spreadsheet. SaaS leaders know this, which is why forecast calls exist — to stress-test the numbers through conversation. But that process is time-consuming, subjective, and still dependent on the quality of the underlying CRM data.
Most mature SaaS teams use a hybrid: stage-based pipeline as the baseline, rep commits as a sanity check, and AI signals to flag deals that are diverging from what the rep is reporting.
The single biggest reason SaaS forecasts miss is bad CRM data. If MEDDPICC fields are empty, close dates are stale, and deal stages haven't been updated since last month's pipeline review, the forecast is fiction. Garbage in, garbage out — no matter how sophisticated your forecasting model.
The second reason is rep optimism. Salespeople are wired to believe their deals will close. Commit calls become exercises in conviction rather than analysis. Without an independent signal — something that doesn't rely on the rep's self-assessment — it's very hard for managers to distinguish a deal that's genuinely on track from one that's wishful thinking.
The third reason is late-stage surprises. A deal that looked solid in week 8 of a 12-week quarter goes dark in week 10. The economic buyer was never actually engaged. Legal has concerns nobody surfaced. A competitor got in. These surprises are usually not surprises at all — the signals were there in the deal activity, the MEDDPICC gaps, the declining email response rates. They just weren't visible in time.
The best SaaS revenue teams treat forecasting as a continuous process, not a weekly ritual. Rather than scrambling to update CRM data before a Thursday forecast call, they maintain pipeline hygiene in real time — so that the forecast at any moment reflects the actual state of the business.
Good forecasting practices include: weekly pipeline reviews that inspect deal health rather than just deal value, a clear definition of what "commit" means and consistent enforcement of it, MEDDPICC as a qualification standard rather than an optional framework, and regular inspection of deal activity signals — not just rep-reported stage.
The shift from reactive forecasting (cleaning up data before a call) to proactive forecasting (maintaining data quality continuously) is the single biggest lever most SaaS teams have for improving forecast accuracy. AI makes this shift possible at scale.
AI changes forecasting in two fundamental ways. First, it ensures the underlying CRM data is accurate by automating the capture of deal information from calls, emails, and meetings — so MEDDPICC fields, close dates, and activity logs reflect reality rather than memory. Second, it adds an independent signal layer that assesses deal health based on behaviour, not rep opinion.
When Brazn analyses a deal, it looks at the full picture: how often is the economic buyer engaging? Are email response times increasing or decreasing? Has the champion gone quiet? Is the decision timeline slipping? These signals, aggregated across all your open opportunities, produce a deal health score that correlates strongly with actual close probability — independent of what the rep has entered in Salesforce.
For CROs and VP Sales, this means the forecast conversation changes. Instead of spending 45 minutes asking "are you sure this will close?", you spend 10 minutes on the two deals where the AI signal and the rep commit are diverging. Everything else you can trust.
See how Brazn improves forecast accuracy for SaaS teams.---
Book a demo to see how Brazn AI fits into your sales stack.
About the Author
Alex Margarit, Sales AI Expert, SaaS Sales Leader, BMC, ServiceNow, Docusign — 25+ years in SaaS sales.