Key Takeaway: Here is how to create a sales forecast. List every open deal due to close in the period. Give each a probability based on what the buyer agreed to. Multiply, and add. Then compare it to what closed. The math takes 8 steps. The hard part is the probability. In Ken Lundin’s experience at RevHeat, a forecast built on what reps hope drifts. One built on buyer evidence holds.
What is a sales forecast?
A sales forecast is an estimate of the revenue your team will close in a set period. The period is usually a month or a quarter. According to Salesforce, HR uses forecasts to plan hiring, and finance uses them to plan investments. It tells leadership what to expect, and it tells the sales leader where to step in before the period ends.
What are the main sales forecasting methods?
| Method | How it works | Good for |
|---|---|---|
| Stage-weighted pipeline | Each pipeline stage gets a win probability. Deal value × stage probability = forecast value | Teams with a defined sales process and clear stage rules |
| Historical run rate | Use last year’s same period, adjusted for growth | Stable businesses with steady deal flow |
| Rep commit (categories) | Reps sort each deal into commit, best case or pipeline | Teams with a small number of large deals |
| Length of sales cycle | Probability rises as a deal approaches your typical cycle length and falls once it runs well past it | Teams with a consistent cycle length |
You can combine them. One setup is a stage-weighted forecast checked against rep commit.
How to create a sales forecast, step by step
- Pick the period. Month or quarter.
- Pull every open deal with a close date in the period. Clean up close dates that have already passed.
- Set stage rules. Write down what has to be true, from the buyer, for a deal to be in each stage.
- Set a probability for each stage. Use your own history: of the deals that reached this stage, what share closed?
- Multiply and add. Deal value × stage probability, summed across deals.
- Sort deals into commit, best case and pipeline.
- Check the evidence on the biggest deals. Listen to the last call. Did the buyer agree to what the stage says?
- Compare to actuals at the end of the period and adjust the stage probabilities.
Sales forecast example
The numbers below are made up for illustration.
Stage probabilities, taken from this team’s own history:
| Stage | Probability |
|---|---|
| Discovery complete | 10% |
| Decision maker met | 25% |
| Proposal reviewed with buyer | 50% |
| Verbal agreement | 80% |
Open deals for the quarter:
| Deal | Value | Stage | Probability | Forecast value |
|---|---|---|---|---|
| A | $120,000 | Verbal agreement | 80% | $96,000 |
| B | $80,000 | Proposal reviewed with buyer | 50% | $40,000 |
| C | $60,000 | Decision maker met | 25% | $15,000 |
| D | $150,000 | Discovery complete | 10% | $15,000 |
| Total | $410,000 | $166,000 |
The weighted forecast is $166,000.
Now check the evidence. On the last call for deal B, the buyer said the CFO hasn’t seen the proposal. That doesn’t match “proposal reviewed with buyer.” Moved back to “decision maker met,” deal B is worth $20,000 in the forecast, and the total drops to $146,000. That one call changed the forecast by $20,000.
How do you calculate sales forecast accuracy?
Compare the forecast to what actually closed:
Forecast accuracy = 1 − (|actual − forecast| ÷ actual)
Example: forecast $146,000, actual $130,000. The miss is $16,000. $16,000 ÷ $130,000 = 12.3%. Accuracy = 87.7%.
Track it every period, by team and by rep. A rep who is always high, or always low, is telling you how they read their deals.
Why are sales forecasts wrong?
- Stages based on seller activity, not buyer action. “Proposal sent” says what the rep did. “Proposal reviewed with the buyer” says what the buyer did.
- Probabilities that were never checked against history. Default CRM percentages are someone else’s numbers.
- Stale close dates. Deals that keep sliding stay in the forecast.
- The rep’s summary instead of the call. A rep’s summary is how they remember the call, not what was said.
The skills behind a reliable forecast are skills you can hear on calls. RevHeat analyzed data from Objective Management Group on 12,745 salespeople. It compared the top 10% with the bottom 10% across 30 sales skills. Some of the widest gaps were in Commitment (476.92%), Qualifying (380.00%) and Sales Process (375.00%). CRM Savvy (283.33%) also showed a large gap. A rep who skips any of those feeds the forecast guesses.
How do you make the forecast more accurate?
- Write stage rules as buyer actions. “Buyer confirmed budget,” not “budget discussed.”
- Require evidence to move a stage. A quote or a moment from the call.
- Review the largest deals on the call, not in the CRM.
- Reset probabilities from your own history every quarter.
- Flag deals with no dated next step. Move them back a stage.
RevHeat calls the weekly version of this check The Enforcement Loop:
- Read every call. Not just the ones you had time for.
- Tell each rep what to ask on the next call. Word for word.
- Remember exactly what you told each of them.
- Watch the next call to see if they did it.
- Catch the slip and fix it before it becomes a habit.
- Roll what works into a sharper playbook. Then start over.
Why is a sales forecast harder to trust as the team grows?
At 5 reps, you can ask each rep about each deal and hear the call yourself. At 60 reps in several countries, the forecast is a number typed in a box. In one office “80% likely” means the rep is sandbagging. In another it means the rep is protecting their job.
The evidence check in the worked example above took one call. Doing it for every deal is a different job. At about 3 calls a rep a week, 20 reps make 728 calls a quarter, about 243 hours of listening. If your forecast still runs on opinion, that isn’t a discipline problem. Nobody had the hours.
When the calls behind the forecast get checked every week, here is what changes by size:
- 5 to 10 reps: By Monday you can say why the last one died.
- 20 to 50 reps: Managers report what happened instead of what they think.
- 60 to 100+ reps: The same word means the same thing in every country, and marketing gets what the buyer said rather than a rep’s memory of it.
FAQ
What is the simplest way to create a sales forecast?
A stage-weighted pipeline. Multiply each open deal’s value by the win probability for its stage, then add them up.
What is a good sales forecast accuracy?
Set your own target and track it every period. The trend matters more than any one number: accuracy should improve as stage rules and probabilities get tighter.
How often should you update a sales forecast?
RevHeat recommends a weekly forecast call, with a full reset of stage probabilities each quarter.
What is the difference between a sales forecast and a pipeline?
The pipeline is every open deal at full value. The forecast is the share of that pipeline you expect to close in the period.
What is commit vs best case in a sales forecast?
Commit is what the rep is confident will close in the period. Best case adds deals that could close if things go well. Pipeline is everything else.
How do you create a sales forecast in Excel or Google Sheets?
Make one row per open deal. Put the deal value in column B and the stage probability in column C. In column D, enter =B2*C2. Sum column D for the forecast. Add a column for the date of the last call, so stale deals stand out.
How do you forecast sales for a new business with no history?
Use a bottom-up estimate: expected meetings × expected conversion to a deal × average deal size. Replace the guesses with your own numbers as soon as you have a quarter of data.
Related Reading
Ken Lundin is CEO of RevHeat, the Sales Enforcement company. Over 25 years he has sold, run sales teams and helped 200+ teams, including 5 unicorns, generate $1.5B in client sales, with two exits ($250M and $30M) and four appearances on the Inc. 500. He is a Forbes Business Council member and the author of Strategic Selling Unleashed, Broken Playbook and The SmartScaling Formula.