Start with a clear forecasting question
A sales forecast is an estimate of what you expect to sell during a future period. For a freelancer or small team, it can help with workload planning, purchasing and deciding when to follow up on new opportunities. It is not a target or a promise. Its value comes from making your assumptions visible so you can revise them as you learn.
First, decide what you want to forecast: revenue, number of sales, or both. Then choose a time period that suits your decisions. Monthly figures may be useful for planning capacity, while weekly figures may suit a business with frequent, smaller transactions. Use the same definition of a sale throughout. For example, do not mix signed contracts with invoices paid unless you deliberately track them as separate measures.
Prepare a consistent sales history
Gather sales records for several comparable periods and put them in a simple table. Include the period, number of sales and sales value. If the business offers substantially different services, consider separate rows for each service. Ten small projects and ten large projects may represent very different amounts of work and revenue.
Check for missing records, duplicates and changes in how sales were recorded. Note unusual events rather than quietly deleting them: a one-off large contract, a month when you could not take new work, or a temporary promotion may all affect the average. If your prices changed, older revenue may be less useful for estimating future revenue than older sales counts. In that case, you could forecast the count and apply a current, reasonable average sale value.
Limited history is still worth using, but it calls for caution. A few months of results may reflect timing or chance more than a stable pattern. Keep a note of how much data you used so that anyone reading the forecast understands its limits.
Calculate a simple baseline
An average gives you a starting point. Suppose a team completed 8, 10, 9, 11, 12 and 10 sales over the last six months. That is 60 sales in total, or an average of 10 per month. If the team expects similar conditions and capacity next month, 10 sales is a straightforward baseline forecast. At an assumed average value of $500 per sale, that would imply $5,000 in sales value.
The baseline is not automatically the best prediction. If recent months better represent the current business, compare the full six-month average with a shorter one. In this example, the last three months average 11 sales. If sales regularly rise or fall at particular times of year, compare similar periods where you have enough history to do so. Do not label a single unusually strong month a seasonal pattern.
Choose a method you can explain and repeat. Write down the periods used, the calculation and why you selected that baseline. A transparent estimate is easier to improve than a precise-looking number whose origin nobody remembers.
Adjust for changes you can identify
Next, ask what will be different during the forecast period. A booked project may limit capacity for new work. A returning customer may have already confirmed an order. A service may be discontinued or its price may change. Adjust the baseline only when you can state the reason and the likely direction of the effect.
Keep confirmed work separate from possible work. A lead who has expressed interest is not the same as a signed agreement, and neither is necessarily the same as cash received. If your forecast includes potential deals, list them separately with their expected timing and a cautious estimate of whether they will proceed. This makes it easier to see how much of the total rests on uncertain opportunities.
It can help to create a low, central and high estimate. The central estimate can reflect your baseline and the changes you consider most likely; the other two can show plausible outcomes if fewer or more sales occur. These are planning scenarios, not claims that sales will fall within a guaranteed range.
Compare the forecast with results and revise it
At the end of each period, record actual sales beside the forecast. If you forecast 10 sales and completed 7, note the difference and ask what caused it. Did fewer enquiries arrive, did deals take longer to close, or was delivery capacity lower than expected? A single miss may not tell you much, but repeated misses in the same direction suggest an assumption needs attention.
Update the forecast on a regular schedule using the newest comparable data. Preserve earlier versions or notes so you can see what changed. As your records improve, you may find it useful to forecast different services separately or track enquiries and sales together. Start with the smallest model that supports a decision, then add detail only when it helps you plan more clearly.
