Many important business decisions (hiring someone new, investing in a new acquisition channel, ordering more stock, committing to a new fixed expense) depend on a question rarely answered with real data: how much are we going to sell over the next few months? Without a reasonably grounded sales forecast, that question gets answered with gut feeling, with last month's result extrapolated with no further analysis, or with plain miscalibrated optimism. The CRM, if used well, already contains much of the information needed to do something far more reliable.
What sales forecasting actually is
Sales forecasting is the exercise of estimating, as precisely as possible, how much the business will bill in a future period, based on objective data: open opportunities in the sales funnel, their estimated probability of closing, historical conversion from previous periods, and the industry's own seasonality. It is not an exact science, but done well it enormously reduces the margin of error compared to estimating purely on intuition.
Why the CRM is the natural data source for this forecast
If the sales team correctly logs their opportunities (estimated amount, current stage, probable close date), the CRM already contains, with no extra effort, the raw material for building a reasonable forecast: adding up the value of opportunities in advanced stages, weighted by their historical probability of closing at that specific stage, gives a far more grounded estimate than any calculation done from scratch in a separate spreadsheet disconnected from the sales team's daily reality.
Weighting by probability, not adding up the raw value of everything
A common mistake when doing a homemade forecast is adding up the total value of every open opportunity, as if they were all certain to close. In practice, a freshly opened opportunity has a much lower probability of closing than one already in final negotiation, and adding them with equal weight generates systematically inflated forecasts. Weighting each opportunity by the historical conversion probability of the stage it is at gives a much more realistic figure, even if it looks less optimistic at first glance.
History as a corrector of individual biases
Each salesperson tends to have their own bias when estimating their opportunities: some are systematically optimistic, others conservative. Comparing individual estimates against the actual closing results of previous months allows that bias to be calibrated over time, adjusting each person's future forecasts based on their real historical pattern, instead of taking each individual estimate at face value with no adjustment.
Seasonality: not every month looks alike
Many businesses have clear seasonal patterns (more sales at certain times of year, predictable dips at others) that a well-built forecast must incorporate, comparing not just against the previous month but against the same period the year before. Ignoring seasonality leads to interpreting a perfectly normal, expected dip as an alarm signal, or a usual seasonal uptick as an exceptional success that actually repeats every year.
Using the forecast for concrete decisions, not just for reporting
The real value of a sales forecast is not in the number itself, but in the decisions it allows to be made with more confidence: whether the coming months' forecast justifies hiring someone new, whether purchasing pace with suppliers should be adjusted, whether it is a good time to invest in an additional acquisition channel or, conversely, whether it is wise to be more cautious with fixed expenses given a weaker-than-usual forecast.
The limits of forecasting: it remains an estimate, not a certainty
However well built, a sales forecast remains a probability-based estimate, not a guarantee. External factors (a market shift, the unexpected loss of an important customer, an industry crisis) can throw the actual outcome off the forecast, and it is worth reviewing and adjusting the forecast regularly, rather than treating it as a fixed number decided once a year and then forgotten.
Distinguishing between optimistic, realistic and pessimistic forecasts
Instead of giving a single forecast number, many businesses benefit from working with three scenarios: a conservative one (only near-certain opportunities), a realistic one (weighted by historical probability), and an optimistic one (if everything currently in motion closes favourably). Working with this range, instead of a single figure, helps make different decisions depending on the risk level each type of decision can bear: committing to a new fixed expense should be based on the conservative scenario, while deciding whether an extra acquisition campaign is worth it can lean on the realistic scenario.
Forecasting at the product or service level, not just overall
Beyond the overall billing forecast, breaking down forecasting by product or service line reveals information the overall figure hides: perhaps the business as a whole is on track, but one specific product is declining while another compensates for that drop, a signal worth attention even when the aggregate result looks healthy.
When to distrust a forecast that seems too good to be true
A forecast showing sustained growth month after month with no correction, especially if it contradicts industry experience or the business's known seasonality, deserves a critical review before being taken at face value. It is often a sign that opportunities are not being updated rigorously enough, leaving opportunities in the system that should really have already been marked as lost.
Involving the sales team in interpretation, not just data entry
A common mistake is treating sales forecasting as a purely management exercise, where the sales team only enters data without ever participating in the final interpretation. Involving the team in reviewing and commenting on the forecast provides qualitative context numbers alone do not capture (why a specific opportunity is stuck, what is really going on with a key customer) and increases the team's commitment to the accuracy of the data they themselves enter.
Forecasting and capacity planning, beyond sales figures
A reliable sales forecast does not just help plan revenue, it also helps plan the operational capacity needed to serve it: if the forecast anticipates a sales peak in a specific period, the business can prepare in advance in terms of staffing, stock or production capacity, instead of reacting late once demand has already arrived and capacity is not ready to absorb it.
Communicating the forecast beyond the sales team, with the right nuance
Sharing the sales forecast with other areas of the business (production, finance, customer service) helps the whole organisation prepare coherently for what is coming, as long as it is communicated with the right nuance: a forecast remains an estimate, not a locked-in commitment, and presenting it without that nuance can create rigid expectations that are hard to manage later if reality diverges.
Frequently asked questions
Do I need an advanced CRM to do sales forecasting?
Not necessarily a high-end one, but you do need one where the team consistently logs the estimated amount, stage, and probable close date for each opportunity, which are the minimum data points needed to build a reasonable forecast.
How often should the sales forecast be updated?
A monthly review is usually a good balance for most businesses, although some with very short sales cycles can benefit from weekly reviews to react faster to deviations.
What margin of error is reasonable to expect in a sales forecast?
It varies a lot by industry and process maturity, but as more history accumulates and close probability by stage gets better calibrated, the margin of error tends to shrink noticeably compared to the first months of implementing forecasting.
How do I correct excessive optimism from some salespeople when estimating their opportunities?
By systematically comparing their past estimates against actual closing results, and using that comparison to adjust the weight given to their future forecasts, instead of taking each individual estimate literally with no filter.
Does forecasting work for businesses with very irregular sales and no clear pattern?
It still adds value, though with more uncertainty. In these cases it is worth leaning more on detailed analysis of specific open opportunities than on general historical patterns, which are less reliable when there is no stable underlying pattern.
Should I share the sales forecast with the whole team or just with leadership?
It depends on the business's culture, but sharing at least a summarised version with the sales team usually helps align expectations and helps each person understand how their own work contributes to the overall forecast result.
Should I work with a single forecast number or several scenarios?
Working with several scenarios (conservative, realistic, optimistic) adds more value than a single number, because it allows making different decisions depending on the risk level each type of decision can reasonably bear.
Why should I break down the forecast by product or business line?
Because the overall figure can hide one product declining while another compensates for that drop, a warning sign worth catching and addressing even when the aggregate result looks healthy at first glance.
Should the sales team be involved in interpreting the forecast, not just entering data?
Yes, their involvement provides qualitative context numbers alone do not capture, and increases their commitment to the accuracy of the data they themselves log into the system.
What else is a sales forecast useful for beyond anticipating revenue?
It also helps plan the operational capacity needed (staffing, stock, production) far enough in advance, instead of reacting late once demand has already arrived and the business's capacity is not ready to absorb it.
Should I share the sales forecast with areas of the business beyond the sales team?
Yes, it helps the whole organisation prepare coherently, as long as it is communicated with the nuance that it remains an estimate, not a locked-in commitment, to avoid rigid expectations that are hard to manage later.