How to Forecast Sales Accurately: Methods & Steps (2026)

Sales forecasting is the process of estimating how much revenue a business is likely to generate over a defined period. A reliable sales forecast is based on more than the amount of pipeline sitting in the CRM. It combines current opportunities, historical conversion rates, deal timing, sales-cycle patterns, and the quality of the underlying data.

But a pipeline can look healthy while containing stale opportunities, unrealistic close dates, or deals that aren’t progressing. Effective sales forecasting requires the right method and a consistent process for validating the assumptions behind it. 

What Is a Sales Forecast?

A sales forecast is an estimate of the revenue a sales team expects to close during a specific period, such as a month or quarter. Unlike a sales target, which reflects what the business wants to achieve, a forecast reflects what the available evidence suggests will actually happen.

For example, if a sales team has $1 million in open opportunities, it does not follow that the team will close $1 million this quarter. Each opportunity has a different likelihood of closing, and some may not close within the forecast period at all.

A useful forecast therefore answers three questions:

• Which opportunities are likely to close?
• How much revenue will they contribute?
• When is that revenue likely to be realized?

Sales Forecasting Methods: 6 Approaches to Know

There is no single forecasting technique that works for every sales organization. The right method depends on the maturity of your sales process, the amount of historical data available, the complexity of your deals, and the reliability of your CRM.

1. Opportunity-Stage Forecasting

Opportunity-stage forecasting applies a historical win probability to each sales stage:

Forecast Revenue = Deal Value × Stage Probability

For example, $200K in Discovery at 20%, $150K in Proposal at 50%, and $100K in Negotiation at 70% produces a $185K forecast. This works well when stages are clearly defined and backed by enough historical data, but it can miss deal-level differences. Two opportunities in the same stage may have very different levels of buyer engagement, budget, or momentum.

2. Historical Forecasting

Historical forecasting uses past sales performance to estimate future revenue:

Forecast Revenue = Previous Period Revenue × (1 + Expected Growth Rate)

For example, $800K in quarterly revenue with 20% expected growth gives a $960K forecast. This works best with several years of consistent data and predictable seasonality, but it may not account for changes in pricing, market conditions, sales capacity, product mix, or pipeline quality.

3. Bottom-Up Forecasting

Bottom-up forecasting builds the forecast from individual opportunities:

Deal Value × Probability of Closing = Expected Revenue

For example, five $100K opportunities at 80%, 80%, 50%, 50%, and 20% probabilities produce a $280K forecast. This approach is useful for B2B teams because it requires managers to assess individual deals, identify risks, and then roll those estimates into a team-level forecast.

4. Intuitive Forecasting

Intuitive forecasting relies on the judgment of sales reps and managers. It can be useful when entering a new market or launching a product with limited historical data, but it is difficult to standardize and can be influenced by individual bias. 

It’s best used as one input rather than the entire forecasting method.

5. Multivariable Forecasting

Multivariable forecasting combines signals such as historical win rates, deal size, sales-cycle length, buyer engagement, stakeholder involvement, rep performance, and seasonality. This is better suited to complex B2B sales where pipeline stage alone doesn’t capture deal health. 

New approaches increasingly combine CRM data with engagement signals to improve forecast accuracy.

6. Sales-Cycle Forecasting

Sales-cycle <internal link to sales cycle blog> forecasting uses the historical time required to convert opportunities into customers.

For example, suppose your company typically closes:

• SMB deals in 60 days
• Mid-market deals in 100 days
• Enterprise deals in 180 days

If an enterprise opportunity entered the pipeline 30 days ago, it may be premature to include it heavily in the current-quarter forecast, even if the rep has marked it as a late-stage opportunity.

This method is especially useful for identifying whether a deal's expected close date is realistic.

How to Create a Sales Forecast in 6 Steps

Knowing the sales forecasting techniques is only part of the process. You also need a repeatable workflow for turning pipeline data into a forecast: 

Step 1: Define the Forecasting Period

Start by deciding what you are forecasting: the current month, quarter, or another defined period.

For most B2B sales teams, monthly and quarterly forecasts are more actionable than long-range forecasts because the further out you go, the more assumptions enter the model.

Step 2: Clean and Review the Pipeline

A forecast built on stale CRM data will be unreliable regardless of how sophisticated the formula is. Before calculating anything, make sure the pipeline you're forecasting from is reliable.

Check for:

• Opportunities with outdated close dates
• Deals that have had no recent buyer activity
• Duplicate opportunities
• Missing deal values
• Opportunities sitting in stages that no longer reflect reality
• Deals that have repeatedly slipped between periods

Step 3: Assign Probability to Each Opportunity

Next, determine the likelihood that each opportunity will close within the forecast period. You can use historical stage conversion rates, rep judgment, or a combination of both.

For example: $150,000 opportunity × 60% probability = $90,000 forecast contribution

Repeat this calculation for every relevant opportunity.

Step 4: Identify Deals That Could Distort the Forecast

This is where a basic weighted pipeline forecast needs human inspection.

A deal might have a 70% stage probability but still be risky because the economic buyer has not been involved, the close date has already slipped twice, or the buyer has stopped responding.

Conversely, a newer opportunity may be more likely to close than its stage suggests because the buying process is moving unusually quickly.

This is also the point where forecast accuracy should be reviewed against previous periods. If your team consistently forecasts $1 million and closes $800,000, that historical variance should influence how you interpret the current number.

Step 5: Account for Historical and External Factors

Your pipeline should not be the only input. For example, if your historical forecast has been 10% higher than actual revenue over the past four quarters, that pattern is worth investigating before accepting the current forecast at face value. 

Consider whether the forecast period includes:

• Seasonal changes in demand
• Pricing changes
• New product launches
• Changes in sales capacity
• Major customer or market events
• Historical forecast bias
• Changes in average sales-cycle length

Step 6: Roll Up the Forecast and Review It Regularly

Once individual opportunities have been assessed, roll them up into team, segment, territory, and company-level forecasts.

Then establish a consistent review cadence. Weekly reviews are useful for active pipeline management, while monthly or quarterly analysis can reveal broader forecasting patterns and systematic bias.

The goal is not to produce one number and leave it untouched. A forecast should change when the evidence changes.

Why Sales Forecasts Miss

A deal can remain marked as "Commit" long after the buyer's behavior has changed. A close date can remain untouched after several missed milestones. A late-stage opportunity can have very little actual buyer engagement.

This is where stale CRM data and optimistic stage assignments create problems. The forecast can look precise while being based on outdated assumptions.

How Kris@Work Helps Improve Forecast Visibility

Kris@Work approaches this problem through Kris Close's Health and Projections and Kris AI CRM Reporting, rather than positioning itself as a standalone forecasting module.

Kris Close grades every deal from A+ to F, with the health grade decomposed into contributing signals. It also provides a per-deal ARR projection based on deal evidence, with conversational and CRM data jointly weighted. 

Kris AI CRM's reporting layer brings together pipeline velocity, stage conversion, and forecast against commit from the live CRM record. 

Instead of relying solely on what a rep entered into a stage or forecast category, sales leaders can inspect the evidence behind the deal's current health and projected value.

Capwise, a financial advisory firm, uses Kris Close's A+ to F deal grades and per-deal ARR in its monthly CXO reviews. Its leadership team now runs those reviews from Kris pipeline and reporting views. 

How to Measure Forecast Accuracy

Once you have a forecasting process, measure how well it performs. A simple forecast accuracy formula is:

Forecast Accuracy = 1 - |Forecast Revenue - Actual Revenue| ÷ Actual Revenue

For example, if your forecast was $1 million and actual revenue was $900,000:

1 - ($100,000 ÷ $900,000) = 88.9%

You should also track forecast bias, which tells you whether your team systematically over-forecasts or under-forecasts.

More importantly, compare accuracy by rep, segment, forecast category, deal size, and time period. This can reveal whether the problem is a particular sales motion or a broader issue with pipeline quality and forecasting discipline.

Conclusion 

An accurate sales forecast starts with reliable pipeline data, realistic deal probabilities, and a consistent process for reviewing what is actually happening in each opportunity. The goal is not to predict the future perfectly, but to make the forecast reflect the reality of your pipeline as closely as possible.

Kris@Work helps sales teams do that by grounding deal health and ARR projections in live deal evidence, while its AI CRM keeps the underlying record current

Start a free Kris@Work trial. No credit card is required. 

FAQs

What is the most accurate sales forecasting method?

There is no universally best method. For established B2B teams, a combination of bottom-up pipeline analysis, historical conversion data, sales-cycle information, and deal-level evidence is generally more useful than relying on a single formula.

How often should sales forecasts be updated?

Most sales teams should review forecasts weekly during an active period, with broader monthly or quarterly analysis to identify trends and forecasting bias.

What is the difference between a sales forecast and a sales target?

A sales target is the revenue a company wants to achieve. A sales forecast is the revenue the business currently expects to achieve based on available evidence.

How can I improve forecast accuracy?

Start with clean CRM data, clearly defined stages, consistent probability rules, realistic close dates, and regular pipeline reviews. Then compare forecasts against actual results and use the gaps to improve your forecasting model over time.