Sales forecasting methods compared: historical, weighted pipeline, forecast categories and more, why forecasts miss, and how to make the inputs trustworthy.

Migaru5 min read

Sales forecasting is estimating how much revenue will close in a period, usually a month or a quarter. The main methods are historical (last period plus growth), weighted pipeline (each deal’s value times its stage’s win probability), forecast categories (each rep sorts deals into commit, best case and pipeline), and analytical models that weigh many signals at once. Most teams use two side by side and look hard at where they disagree.

Every method is only as good as its inputs. A forecast misses far more often because close dates and stages were wrong than because the method was. So this guide covers both: the methods, and how to keep what they’re built from honest.

The main sales forecasting methods

  • Historical forecasting. Last period’s revenue, adjusted for growth and seasonality. Simple and useful as a sanity check, but blind to what is actually in the pipeline.
  • Weighted pipeline (stage-based). Each open deal’s value times the win probability of its stage, summed for deals closing in the period. Objective and easy to automate, and only as good as the stages and probabilities behind it.
  • Forecast categories. Each rep places each deal in a category, such as commit (will close), best case (could close) and pipeline (won’t this period), and managers roll the categories up. It captures what reps know that the CRM doesn’t, and also their optimism.
  • Length of sales cycle. A deal’s chance of closing is judged by how far through a typical cycle it is. Useful where cycles are consistent.
  • Multivariable or AI models. Weigh stage, activity, deal age, past slips and more. Strongest with plenty of history, and hardest to explain when they disagree with a rep.

Which method to use

  • Under a few dozen deals a quarter: go deal by deal. Judge each one on evidence, and total the ones you’d bet on. Any formula over so few deals is noise.
  • A small team with a stage-based pipeline: weighted pipeline as the baseline, with a deal-by-deal review of the largest deals.
  • A team with several reps and managers: forecast categories for the call you make, and weighted pipeline as a check on it.
  • Years of clean history: a model can earn its place, as one more check, not as the answer.

Whichever you pick, compare the forecast with what actually closed every period, and write down why each miss happened. That record improves a forecast faster than any change of method.

How weighted pipeline works

Take every open deal with a close date in the period. Multiply each deal’s value by the win probability of its stage, and add them up. A $40,000 deal in a stage that wins 25% of the time counts as $10,000; a $20,000 deal in a stage that wins 75% counts as $15,000.

Set the probabilities from your own history once you have some: of the deals that reached each stage, how many were won? Until you do, keep them rough, and treat the weighted total as a range rather than a number. Sales pipeline stages covers the exit criteria that make those probabilities mean something.

Why sales forecasts miss

  • Close dates that are wishes. A date chosen because it is the end of the quarter, not because the buyer named it.
  • Stages that don’t reflect the buyer. A deal in proposal because a proposal was sent, not because anyone read it.
  • Stale deals. Deals with no conversation in weeks still counted at full weight.
  • Deals pushed before. A deal whose close date has already moved once is likelier to move again, whatever its stage says.
  • Happy ears and sandbagging. Some reps hear yes in every maybe; others keep sure deals out of commit to beat their number.
  • Missing steps after yes. Procurement, legal review and a purchase order take time a verbal yes doesn’t account for.

How to build a forecast you can believe

  • Make close dates the buyer’s. Every date should come with a reason: a budget cycle, a launch, a contract ending.
  • Move stages on exit criteria, and let a deal move back a stage when the evidence says so.
  • Discount deals without recent contact. If nobody has spoken to the buyer in three weeks, it is not closing this month.
  • Track pushes. Record every time a close date moves or a deal goes back a stage, and treat repeat offenders as at risk.
  • Separate commit from the rest. Give leadership the total you would bet on, and the total of everything, and say which is which.
  • Review slips every period. For each deal that was due and didn’t close, find out why. The reasons repeat.
  • Keep activity captured automatically, so “last contact” is a fact rather than whatever someone logged.

CRM hygiene is the groundwork: a forecast built on stale records inherits every one of their errors.

What Migaru does for the forecast

Migaru isn’t a forecasting tool with categories, rep calls and roll-ups. What it does is keep the inputs honest, which is where most forecasts go wrong. Email, meetings and calls are captured on each deal by themselves, so its last and next interaction are always current. Stage changes are suggested from what was said, with a reason, rather than made silently or left to drift.

Each deal’s weighted value is its total value times its stage’s win probability, beside its committed value: what the customer is bound to pay once they sign, kept apart from the months a customer on a rolling contract is expected to stay. Both sort, filter and chart like any field. The agent counts in the database rather than skimming, can rebuild the pipeline as it stood on any past date since your deals came into Migaru, and comes with a Forecast check skill that sorts the deals due in a period into likely, at risk and not this period, treating a deal already pushed once as not likely, and gives the likely total separately.

On Pro and above, What slipped shows what became of every deal that was due to close, with a reason for each one that slipped, and dashboards keep the numbers you check each week. The AI sales assistant covers the rest, and plans and prices what each plan includes.

Frequently asked questions

What is sales forecasting?

Estimating how much revenue will close in a coming period, usually a month or quarter, from the deals in the pipeline, past results, or both. A good forecast says what it is based on and how sure it is.

What is the most accurate sales forecasting method?

None is reliably most accurate on its own. Teams usually get closest by combining a weighted pipeline, which is objective, with a deal-by-deal judgement of the largest deals, and by reviewing why each past forecast missed.

What is a weighted pipeline?

The sum of each open deal’s value multiplied by the win probability of its stage. A $40,000 deal in a stage that wins a quarter of the time counts as $10,000.

What are commit, best case and pipeline?

Common forecast categories. Commit is what a rep is confident will close in the period, best case is what could close if things go well, and pipeline is everything else open. Leadership usually plans on commit.

Why are sales forecasts often wrong?

Mostly because of their inputs: close dates that don’t come from the buyer, stages that don’t reflect what the buyer has done, deals with no recent contact counted at full weight, and deals that have slipped before treated as if they won’t again.

See what Migaru does with your customer record.