Why sales forecasting matters, and how to do it well

Sales forecasting is estimating how much revenue your company will bring in over a set period, such as the next month or quarter. It matters because almost every other plan depends on it: how many people you hire, how much you spend on marketing, how much cash you will have and which targets you set. A good forecast also shows trouble early, while you still have time to fix it.

Why is sales forecasting important?

It turns goals into numbers

A forecast forces you to answer specific questions. How many new customers do you expect this quarter? What will each one spend? How much will existing customers expand or churn? For subscription businesses, that usually means forecasting monthly recurring revenue, not just new deals.

It lets you plan spending and hiring

If the forecast shows you will beat target, you can hire, buy stock or invest in a new channel with more confidence. If it shows a gap, you can act now: push more effort into lead generation, focus reps on expansion deals, or delay spending. Comparing actual results with the forecast each week shows early whether the plan is on track.

It builds trust with investors and the board

Investors want to see a clear path for revenue and evidence that you understand your own numbers. A forecast that you can explain, and that you hit consistently, makes every other conversation easier.

It exposes weak spots in the sales process

When you forecast deal by deal, you see where deals stall. If many deals sit in the proposal stage for weeks, that stage needs attention. Over time, forecasting shows which stages, reps or deal types are hard to predict.

It keeps the team aligned

A shared forecast gives sales, marketing, finance and leadership the same picture. When the outlook changes, everyone hears it at the same time, along with the plan to respond. Our guide on how to align your sales and marketing covers this in more depth.

What are the main sales forecasting methods?

Top-down forecasting

Leaders start from the total addressable market, estimate the share the company can win and turn that into a revenue figure. It is quick and useful for long-term planning or a new market, but it ignores what is actually in the pipeline.

Bottom-up forecasting

Reps estimate what they will close from their own accounts and opportunities, and managers roll those numbers up. It uses the people closest to the deals, but it depends heavily on their honesty and judgement.

Weighted pipeline forecasting

Each open deal's value is multiplied by its probability of closing, usually set per pipeline stage, and the results are added up. A $10,000 deal at a 40% stage counts as $4,000. CRMs calculate this automatically. HubSpot, for example, calls it the weighted amount: the deal amount multiplied by the deal probability. Many teams also sort deals into forecast categories such as pipeline, best case and commit. Compare tools in our list of the best CRM systems for SMBs.

Time series and other statistical methods

Statistical methods use past sales to project the future, for example with moving averages, trend and seasonal models, or regression on drivers such as marketing spend. They need enough history and work best when the future looks like the past.

Judgement-based methods

When there is no history, such as for a new product, forecasts rely on judgement: executive opinion, a sales force composite, customer surveys, forecasting by analogy with similar launches, or the Delphi method, where a panel of experts forecasts anonymously over several rounds.

Which method is most accurate?

There is no single winner. The forecasting textbook Forecasting: Principles and Practice by Rob Hyndman and George Athanasopoulos notes that statistical forecasts are generally better than forecasts based only on judgement, and that judgement should mainly add information a model doesn't have. For a B2B team, a practical setup is a weighted pipeline plus documented rep judgement for the current quarter, and a statistical model for longer periods.

What affects the accuracy of a sales forecast?

How predictable something is depends on how well you understand what drives it, how much data you have, and how similar the future will be to the past. In sales, these factors move the numbers most often:

• Market conditions: economic shifts, new competitors, or a rival's big campaign or price cut.
• Product changes: a new feature can lift sales; a new product can cannibalise an old one.
• Seasonality: budget cycles, holidays and quarter-end buying patterns.
• Marketing activity: campaigns, channel changes and budget shifts. See our guide to multi-channel marketing.
• Regulation: new rules can open or close a market.
• Team size and tenure: new reps need time to ramp; departures leave pipeline uncovered.
• Internal policies: a change to commission plans or discount rules changes when deals close.
• Territory changes: reps moving to new accounts usually close less while they learn them.

What goes wrong with sales forecasts?

The textbook lists familiar problems with judgement-based forecasts:

• Optimism: people forecast what they hope for, and group meetings can amplify it.
• Sandbagging: when the forecast is also the target, a rep may forecast low to beat it.
• Anchoring: forecasts stick close to last quarter's number and are slow to react to new information.
• Inconsistency: the same person forecasts differently depending on mood or recent events.

The cost of a bad forecast runs both ways. Overestimate, and you overspend, over-hire and may have to cut later. Underestimate, and you miss the chance to hire or invest in time.

How do you make sales forecasts more accurate?

1. Keep CRM data clean. Every deal needs a realistic amount, stage and close date.
2. Define your stages and probabilities based on what has actually closed before, not on guesses.
3. Separate forecasts from targets. The forecast is what you expect; the target is what you want.
4. Document adjustments. If a manager changes a number, write down why.
5. Measure accuracy every period. Compare forecast with actuals using a simple measure such as the average absolute error or the average percentage error, and look for patterns.
6. Review regularly. A short weekly review of changes in the forecast catches problems while there is still time to act.

For the metrics that feed a good forecast, see our guide to SDR KPIs and why tracking leads matters.