How to Forecast Sales With Only 20 Deals in Your Pipeline

If you have about 20 open deals, forecast them one at a time and give yourself a range instead of one number. Sort each deal into committed, likely or possible, based on what the customer has actually said or done. Then total each bucket. That takes less than an hour, and it's more honest than multiplying a pile of deals by a stage percentage.

This article explains why the usual method wobbles on a small pipeline, and how to build a three-number forecast you can trust. It ends with a checklist you can run every week.

Why the usual percentage forecast wobbles at small scale

Most forecasting guides teach the weighted pipeline: multiply each deal's value by the close probability of its stage, then add everything up. The most widely used method is the weighted forecast: multiply the value of each deal by the close probability of the stage it is in, and add it up. It works well when you have lots of deals.

The trouble is that it relies on averages, and averages need volume. Weighted forecasting assumes that across many deals, the ones that beat their probability offset the ones that miss. One small-team guide points out that across hundreds of deals those errors cancel out, but across the twenty a small team closes in a quarter they do not, so the forecast is unbiased on average and wrong almost every time.

There are two other problems.

• A deal is all or nothing. A deal closes at full value or at zero, and with only ten open deals one deal landing a month late changes the whole month. A "$6,000 × 50% = $3,000" line doesn't match anything that can happen.
• Your own win rate is shaky. With three sales representatives and twenty deals a quarter, one close moves your win rate by five points, so any statistical method looks unstable.

So percentages aren't useless, but a single weighted total can look precise when it isn't. Treat it as a cross-check, not the answer.

When you may not need a forecast at all

Sometimes the simple answer is the right one. According to one guide, a small team can skip forecasting in two situations:

• You have fewer than five open deals. You can hold five deals in your head, and percentages on five deals tell you nothing you do not already know.
• Your sales close within a week. If people buy in days, last month's sales are a better guide to next month than any list of open deals.

If neither describes you, because you have somewhere between about 6 and 40 deals that take weeks or months to close, read on.

Step 1: Clean the pipeline first

A forecast is only as good as the deals in it. Before you add anything up, fix the two inputs that cause the most damage. One source describes them as expected close dates that nobody updates, and stage definitions that describe your activity rather than the buyer's.

Go through every open deal and ask:

1. Is it still alive? Deals that have quietly died tend to stay open. Closing them as lost feels like admitting defeat, so they linger in the pipeline inflating the total. Close them as lost. You can reopen one if the customer comes back.
2. Is the close date realistic? Move any date that has already passed, and set it to when the customer says they will decide, not when you hope they will.
3. Does the stage describe the buyer, not you? "Proposal sent" is something you did. "Customer has reviewed the proposal and asked about start dates" is something they did. Write a one-line exit rule for each stage so a deal can't move forward on hope.
4. Has it gone quiet? A deal that has sat in "proposal" for six weeks is not a live proposal, it is a stalled one, and treating it as forecastable is how misses happen. Your own sales cycle sets the right threshold. Six weeks is just one example.

This step shrinks most pipelines, and that's fine. A smaller, real pipeline beats a bigger, imaginary one.

Step 2: Sort deals into committed, likely and possible

With a small pipeline, forecast deal by deal rather than by applying probability percentages to a total, sorting open deals into committed, likely, and possible using evidence from the customer.

Here is a practical way to define the buckets. Adjust the wording to fit your business.

Committed

The customer has said yes, or has done something that only a buyer does. Examples are a signed quote, a deposit, a confirmed start date, or a purchase order on the way. You're waiting on paperwork, not on a decision.

Likely

The customer has a real need, a budget and a decision-maker involved, and the next step is agreed and dated. They have seen your price and haven't objected. You'd be surprised, not shocked, if it didn't close.

Possible

Everything else that's still alive. The customer is interested, but you're missing something: budget, timing, the decision-maker, or a reply since your last message. These deals can close. The point is that forecasting them is guesswork, so their value belongs in your upside, not in your number. As one guide puts it, their upside belongs in the range rather than the number.

The rule that keeps this honest is that evidence from the customer decides the bucket, not your gut feeling about the customer.

Step 3: Turn the buckets into three numbers

The advice from the same guide is to give three numbers rather than one, and be explicit about what each assumes. Here is a worked example. The figures are made up to show the arithmetic, so they are not benchmarks.

Say your target for the quarter is $60,000, and you have 12 open deals:

The three numbers are:

• Low case: $20,000. Only the committed deals close.
• Expected case: $32,000. Committed deals plus your own close rate on the likely bucket. If you've historically closed 40% of likely deals, that's $20,000 + ($30,000 × 0.40) = $32,000.
• High case: $50,000. Committed and likely deals all close. The possible bucket stays out. If any of it lands, it's a bonus.

Against a $60,000 target, even the high case falls short. That's useful to know in week two of the quarter, because you can still do something about it.

Where does your own close rate come from?

Use your own history, not an industry figure. For the expected figure, use your own conversion rate on the likely bucket rather than a generic percentage. After two or three quarters you will know it: some teams close 60% of likely, others 35%.

Until then, count your recent closed deals, won and lost. One small-team guide suggests that with fewer than ten closed deals you do not have enough history to count. In that case, use a cautious guess for the likely bucket. Write it down so you can check it against reality at the end of the quarter, and replace it once you have some history.

Step 4: Sanity-check with your run rate

Your deal-by-deal forecast shouldn't float free of your track record. One source recommends that small teams rely on sales judgment structured as commit, best case and pipeline, and use run rate as the sanity check.

Look at what the last few quarters actually brought in. If you closed roughly $30,000 to $35,000 each of the last four quarters and your expected case is $32,000, that agrees. If your expected case is $55,000 and nothing has changed in your business, ask what's different, and find the specific deals that justify it.

Step 5: Check whether you have enough pipeline

The forecast tells you what's likely to close. A separate question is whether you have enough deals in the funnel to reach your target. That's your pipeline coverage ratio.

The common "3x" rule isn't universal. A healthy ratio is 1 divided by your win rate, so a team that wins 25% of its pipeline needs 4x. 3x is enough only if you win about one in three of your deals. With a 20% win rate you need 5x.

If the number is too low, you have a lead-generation job to do, not a forecasting one. Do it early in the period, while new deals still have time to mature.

Step 6: Review weekly and track how long deals sit

A forecast goes stale quickly. A forecast from three weeks ago describes a pipeline that no longer exists. At the same time, don't overreact to every small move. One source warns that forecasting too often on a small deal count produces motion rather than information. A weekly review of bucket changes is plenty. You'll find that most of your attention goes to the few deals that decide the quarter.

Once you have enough closed deals, add one more habit: note how long deals typically sit in each stage. Once you have thirty or forty closed deals, you know your median time in each stage, and a deal well past that median is not on track, whatever bucket it sits in.

As for accuracy, one guide suggests aiming for within about 20% at the start of a quarter and considerably tighter by the final month. That's a reasonable target, and one source warns that chasing more precision early on gives you false confidence.

Your weekly forecasting checklist

Run through this once a week. It takes 20 to 30 minutes.

1. Close any deal that's dead as lost, with a short reason.
2. Update expected close dates, with no dates in the past.
3. Move each deal to the bucket the customer's latest actions justify: committed, likely or possible.
4. Write down the next step and its date for every committed and likely deal.
5. Total the buckets into your low, expected and high numbers.
6. Compare the expected number with your run rate and your target.
7. Check coverage: is there enough live pipeline for the gap?
8. Pick the two or three deals that will make or break the quarter, and agree the next step on each. This habit comes from identifying the two or three deals that will make or break the quarter and agreeing the next step on each.
9. At quarter end, compare your forecast with the result, and adjust your likely-bucket close rate.

Common mistakes to avoid

• Counting hope as evidence. "They loved the demo" isn't a buying signal. A dated next step is.
• Reporting one precise-looking number. Three numbers with stated assumptions are more useful, because they show how much depends on a few deals.
• Leaving stalled deals at full weight. Stalled deals are the single biggest source of weighted-pipeline inflation.
• Borrowing someone else's win rate. Your prices, buyers and sales cycle are your own. Use your numbers.
• Buying complexity too early. Teams with fewer than 200–300 closed deals in Salesforce don't have enough history for ML models to outperform simpler rule-based scoring. A spreadsheet or a simple CRM board is enough for now.

Making the weekly review easier

The weekly review is only as quick as your data is tidy. If your deals, tasks and customer notes sit in one place, the bucket sort takes minutes. In Salestead CRM, the deal pipeline sits alongside tasks and invoices, and the built-in assistant, Lime, can answer questions about your pipeline, forecast revenue, and flag deals at risk in a daily briefing. Treat what it gives you the same way as the rest of this article: as a second opinion to check against your own deal-by-deal view.

If you outgrow the three-bucket method, that's a good problem. As your closed-deal history builds, your stage close rates become reliable enough to trust a weighted forecast, and you can run it next to the range.

Try it free

If you'd like to run this process in a CRM built for small businesses, you can try Salestead free at salesteadcrm.com. The Free plan costs $0, so you can set up your pipeline, sort your deals and see how your next quarter looks before you spend anything.