Sales forecasting techniques are structured methods businesses use to predict future revenue from pipeline data and past sales results. The five most common are historical, pipeline, opportunity stage, bottom-up and multi-variable analysis.
No single technique works for every team.The best technique depends on your team’s size, your sales cycle length and how much historical data you have on hand. This guide compares all five methods, their strengths, their limits and when to use each one, to help you find the right approach for how your team sells.
If you’re new to the topic, our sales forecasting 101 guide 101 guide covers the fundamentals – it’s a good idea to read that before you compare the five techniques below.
Key takeaways
Sales forecasting techniques use different inputs and suit different conditions. Historical forecasting draws on past results, while pipeline and opportunity-stage methods rely on current deal data.
Choose a technique based on the quality of your historical and pipeline data, sales-cycle complexity, team resources and the level of detail decisionmakers need.
Combining techniques can help teams balance historical trends with current pipeline signals, provided each input is reliable and the calculation is applied consistently.
Start with the method your existing data can support, then compare forecasts with actual results and refine the model as your data quality improves.
Why the forecasting technique you choose matters
A forecast only helps if people trust the numbers. If you pick a technique that outpaces a team’s data or sales process, the output will look precise while hiding a guess underneath.
Say a five-person team without much sales history tries multi-variable analysis forecasting. The model needs consistent data across several sources and thin history turns the weighting into a guess. The forecast ends up looking sophisticated but is less reliable than a simple historical average.
The reverse problem hits larger teams too. A 200-person sales org running bottom-up forecasting alone asks a manager to sanity-check hundreds of individual rep estimates by hand every cycle - a task breaking down long before a team grows this large.
Once a forecast stops matching what closes, sales leaders stop trusting the number. Reps sandbag their estimates, finance stops planning around the figure and the forecast turns into a report nobody reads.
Matching a technique to a team’s size, sales cycle and data on hand is what keeps a forecast worth building in the first place. The five techniques below solve this matching problem differently, each fitting a different combination of team size, data and sales cycle.
5 sales forecasting techniques, compared
Each of these sales forecasting methods draws on a different mix of pipeline data, past performance and rep input. The breakdown below covers the data each technique needs, the gaps in each one and which sales teams tend to benefit most.
Historical forecasting
Historical forecasting takes what happened in your past sales cycles and projects the results forward. If your team closed 20% of qualified deals last quarter, this method assumes a similar close rate going forward and applies the rate to your current pipeline value.
Say your team has 40 qualified deals worth $10,000 each in the pipeline this quarter. At a 20% historical close rate, historical forecasting would put your forecast at roughly $80,000.
This method needs at least a few quarters of clean historical data to be reliable. Historical forecasting works best for established teams selling a stable product to a familiar market.
The main limitation comes from assuming the future looks like the past, which makes historical forecasting weak for new products, new markets or any team going through fast change.
Pipeline forecasting
Pipeline forecasting treats your whole pipeline as one data set instead of scoring each deal individually. A historical conversion rate for each stage, built from hundreds of past deals, applies to the total value sitting in that stage today.
Opportunity stage forecasting, covered next, assigns a probability to each individual deal instead, so the two methods work at a different level of detail.
Say $500,000 in pipeline value sits in your negotiation stage. Deals at that stage have historically closed 60% of the time, so pipeline forecasting counts $300,000 toward your total forecast from this stage alone. Repeat the calculation across every stage and add up the results.
This method needs a solid volume of historical deals, often hundreds per stage, to calculate a conversion rate worth trusting. Pipeline forecasting works best for teams with consistent pipeline management and a steady flow of deals each quarter.
The main limitation comes from relying on averages, since a stage full of unusually large or unusually small deals skews the rate for the whole group.
Opportunity stage forecasting
Opportunity stage forecasting assigns a probability to each individual deal, based on which stage that deal sits in right now. Unlike pipeline forecasting, which calculates an aggregate rate from historical data, these probabilities usually come from a rule of thumb or a default set in your CRM.
Say a $50,000 deal sits in your proposal stage and your team assigns proposal-stage deals a 50% probability of closing. Opportunity stage forecasting counts $25,000 of that deal toward your forecast. Add up every deal in your pipeline the same way to get a total number.
.
Note: In Pipedrive, a deal-specific probability overrides the stage’s default probability once you set it and it still counts toward that stage’s overall weighted value.
This method needs almost no historical data to get set up, since a probability gets assigned by stage rather than calculated from past results. Opportunity stage forecasting works best for teams that want a quick, easy-to-explain forecast for leadership without building anything complex.
The main limitation sits in how the probabilities get set. Sales ops or a CRM default assigns them, often from a general sense of how deals move rather than rigorous analysis, so opportunity stage forecasting sometimes overstates confidence in the number.
Bottom-up forecasting
Bottom-up forecasting starts with individual reps and works upward. Each rep estimates what they expect to close from their own deals and those individual estimates add together into one team forecast, rather than starting from a top-down revenue target.
Say five reps each estimate they will close $40,000 this quarter based on the deals in their own pipeline. Bottom-up forecasting adds those five estimates together for a team forecast of $200,000, built from the ground up rather than handed down as a quota.
This method needs reps who track their own deals closely and take the time to give honest, deal-level estimates each cycle rather than rough guesses. Bottom-up forecasting works best for smaller teams where a sales manager knows each rep’s pipeline well enough to sanity-check the numbers.
The main limitation comes from relying on individual judgment since a rep who feels optimistic, or one who sandbags to protect against missing a future target, skews the total in either direction.
Multi-variable analysis forecasting
Multi-variable analysis forecasting blends several types of data into one weighted model, including historical close rates, current pipeline data and outside market factors like seasonality or industry trends. Combining several inputs this way reduces how much a single wrong input skews the final number.
Say historical data points to a 70% close rate, pipeline data suggests 65% given current stage mix and a seasonal slowdown is expected next month. Multi-variable analysis forecasting weighs all three inputs together instead of relying on a single input, landing on a blended figure that accounts for each factor.
This method needs mature reporting tools and enough data across several sources to build a reliable model, which usually rules out small or newly formed sales teams. Multi-variable analysis forecasting works best for larger, established teams with the CRM analytics and reporting infrastructure to support the approach.
The main limitation comes from complexity, since building and maintaining a multi-variable model takes more setup work than the other four techniques and a poorly weighted model sometimes misleads a team despite pulling from more data sources.
Technique | Best fit | Data needed | Effort | Main limitation |
Historical forecasting | Established teams in a stable, familiar market | A few quarters of clean historical data | Low | Assumes the future mirrors the past |
Pipeline forecasting | Teams with a consistent, high-volume sales process | Historical conversion rates for each pipeline stage | Moderate | Averages miss unusually large or small deals in a stage |
Opportunity stage forecasting | Teams that want a fast, easy-to-explain forecast | Minimal, mainly assigned stage probabilities | Low | Probabilities are often guessed rather than data-derived |
Bottom-up forecasting | Smaller teams with close manager visibility into each rep’s pipeline | Honest, deal-level estimates from every rep | High | Vulnerable to individual optimism or sandbagging |
Multi-variable analysis forecasting | Larger, established teams with mature reporting tools | Historical, pipeline and external market data combined | High | Complex to build and a poorly weighted model sometimes misleads |
How to choose the right sales forecasting technique for your team
Team size, sales cycle length, how much clean historical data you have and how much time your team has to maintain the forecast each cycle will determine the right technique for you. Weigh all four before you pick one.
Team size affects how much effort a technique takes to run well. A five-person team lets a manager sanity-check each rep’s estimate by hand, so bottom-up forecasting fits a team this size.
A 200-person sales org needs pipeline or multi-variable forecasting instead, since no manager reviews hundreds of individual estimates every week.
Sales cycle length affects how fast reliable data builds up. Short cycles close enough deals each quarter to make historical and pipeline forecasting practical within a few reporting periods. Long cycles close fewer deals per period, so opportunity stage or bottom-up forecasting often fit better until historical data accumulates.
Historical data quality decides whether the math holds up. Clean, multi-quarter data supports historical or pipeline forecasting. Thin or newly tracked data points toward opportunity stage forecasting, since this method needs no historical baseline to start.
Maintenance time separates the low-effort methods from the high-effort ones. Opportunity stage and historical forecasting take very little upkeep once set up. Multi-variable and bottom-up forecasting both demand regular attention, whether this means updating a weighted model or checking in with reps each cycle.
Combining two of these techniques often works well once a team’s data and processes support the extra step, rather than relying on a single method. Pipedrive’s sales forecasting software and insights and reports both make tracking whichever technique or blend you choose easier.
Pipedrive in Action: Canine Protection International adds $50,000 in six months by fixing its forecasting process. Canine Protection International, a Dallas-based security company, called revenue forecasting a stumbling block for years. Once the team started weighing close data, deal value and probability together instead of guessing revenue each month, results changed fast. Revenue grew by $50,000 within six months of making the switch.
Read more: https://www.pipedrive.com/en/case-studies/canine-protection-international-case-study
Final thoughts
There’s no single best sales forecasting technique. Historical, pipeline, opportunity stage, bottom-up and multi-variable analysis each serve a different team and the right pick depends on your team size, sales cycle and data on hand. Blending two of these methods is worth considering once your data and process can support the extra step.
Try Pipedrive for free. No credit card required.
