Sales forecasting uses historical performance, pipeline data, and buyer insights to predict future sales. The four main sales forecasting techniques are historical, weighted pipeline, bottom-up, and funnel forecasting. Together, they help businesses estimate which deals will close, when, and how much revenue to expect.
Predicting future sales sounds simple on paper.
You look at what's in your pipeline, estimate which deals will close, and add up the numbers.
But sometimes (ok, a lot of times) deals get delayed and budgets change.
Because you’re never going to be able to 100% prevent all of that, the trick is knowing how to account for it without just guessing.
In this guide, we'll break down how to forecast sales, which techniques work best, and how to use real data and buyer behavior to build forecasts you can trust.
Sales forecasting is an estimate of how much revenue your business will bring in over a set period, usually a month or a quarter. Teams build it from past sales results and the deals currently open, then adjust for how often deals like those have closed before.
A good forecast covers when a deal will close and when the money will show up, because those two dates can be far apart. A customer might be fully ready to buy, but if legal pushes the signature into next quarter, the revenue lands later than you planned.
A forecast lets you plan around the revenue you’ll realistically get instead of the revenue you’re just hoping for. And getting that number right isn't getting any easier as 69% of sales-operations leaders say forecasting is harder than it was three years ago.
A good forecast also gives you an early warning when something’s off. You’d much rather learn that now than on the last day of the quarter.
A target is what you want to hit, while a forecast is what your pipeline and win rates say you’ll probably hit.
It’s an easy mistake to keep nudging the forecast up to match the target, but keeping the two separate is what makes the forecast useful, since the target is where you’re aiming and the forecast is where you’re headed if nothing changes.
The four main sales forecasting techniques are:
Historical and funnel forecasting work well for planning ahead, while weighted pipeline and bottom-up forecasting give you a better view of what your current opportunities will probably deliver.
Historical sales forecasting uses past performance to estimate what a business will sell in the future. If you know how your sales team has done before, you at least have a starting point for what comes next.
It looks at patterns in your previous results to set expectations for an upcoming period. The data that tends to help most includes:
You can also break the numbers down by lead source or service line, which matters because different types of opportunities often have very different win rates and sales cycles.
One simple approach is to average your sales over comparable periods.
It works best when you need a baseline for revenue planning, since it shows what your business has typically been able to generate and whether your current expectations are reasonable. What it can’t do is tell you whether the deals in your pipeline will close, so it pairs best with techniques that look at current opportunities and buyer activity.
Weighted pipeline forecasting estimates future sales by multiplying the value of each open opportunity by its probability of closing. Instead of assuming every deal in your pipeline will close, you give each one a percentage based on how likely it is to become a customer, which gives you the expected value of your current pipeline.
If you have a $50,000 opportunity with a 60% chance of closing, you wouldn’t count the full $50,000 toward your forecast. You’d count $30,000, then repeat that for every opportunity and add up the results.
The tricky part is where the probabilities come from. Some teams use default percentages tied to deal stages, which is fine as a starting point but shouldn’t replace a real look at the opportunity.
To get the total, add up the weighted values of all the relevant opportunities.
A probability is only as good as the evidence behind it, so if your CRM automatically assigns 80% to every deal in negotiation, that doesn’t mean each of those deals has an 80% shot.
Weighted forecasts can also hide how much you’re leaning on one big opportunity. Running scenarios with and without your biggest deals is better than relying on a single weighted number.
Weighted forecasting also says nothing about when a deal will close or when the revenue will start, so you’ll need to answer those questions separately.
Bottom-up forecasting starts with individual opportunities instead of a revenue target or a historical average. Sales managers go through deals with their reps, judge how likely each one is to close, and roll those judgments up into a broader forecast.
It means reviewing each deal and deciding how much confidence the team should have in it, which takes more than looking at the amount and the CRM stage. For each opportunity, the team should know whether the buyer is likely to move forward and whether the deal can realistically close within the forecast period. They should also know which approvals or decisions are still outstanding, what could delay or kill the purchase, and who owns the next action and by when.
The most common categories are:
A category isn’t proof that a deal will close. A rep might feel great about a deal because the buyer has been enthusiastic from day one, but if that buyer still needs budget approval and a legal review, plus a signature from someone who hasn’t been involved yet, plenty can still go sideways.
It’s most valuable when you’re predicting sales for the current month or quarter. At that point historical averages won’t cut it, and you need to know which specific deals can realistically close before the period ends. It also gives managers a chance to spot risks and work with reps on what could move a deal forward.
Funnel-based forecasting estimates future bookings by looking at how leads move through your sales process and how many end up as customers. Rather than starting with existing deals, it works backward from lead volume, conversion rates, win rates, and average deal value. That makes it especially good for figuring out whether you’re generating enough future pipeline to support your sales goals.
It follows the path from lead generation to closed-won business.
Apply your historical conversion rates at each stage and you can estimate how much business your current lead volume could eventually produce. Not every lead deserves equal treatment, though. A partner referral may qualify at a different rate, win at a different rate, or take longer to close than a marketing-generated inquiry, and using one conversion rate across every source makes your forecast look more precise than it really is.
If your average sales cycle runs several months, leads that show up near the end of the quarter may not become customers until the next quarter or later, so funnel forecasting has to account for timing as well as conversion rates.
You also need consistent qualification definitions and a way to avoid double counting. Once a lead becomes an opportunity, it shouldn’t show up in the same forecast as both a lead-based projection and a separate opportunity.
It’s best for longer-term revenue planning, since it shows whether your lead generation and qualification efforts are producing enough opportunities to support future sales. It can also get Marketing and Sales working from the same assumptions about lead quality, conversion rates, and expected pipeline.
For near-term forecasting, funnel projections are less concrete than reviewing real qualified opportunities. That’s why funnel forecasting works best next to historical performance and deal-level forecasting instead of replacing them.
You need a consistent process for checking your data against your pipeline and deciding which deals will likely close.
Here’s a good plan:
Decide whether you’re forecasting the next month, the next quarter, or something longer, and agree on which revenue metric you’re measuring.
The three common options are:
These numbers can’t be swapped for each other. A customer might sign a $60,000 annual agreement in March but not start receiving services until April, so that deal counts toward March bookings and adds nothing to March recognized revenue.
And when teams aren't on the same page, forecasts suffer as three in 10 leaders attribute missed forecasts to weak collaboration between teams. Make sure everyone is forecasting the same thing, because otherwise you’ll end up with several numbers that are all technically correct and answer totally different questions.
Next, look at how your team has done in the past. History gives you a baseline for what you can reasonably expect, which beats picking a revenue goal and working backward to justify it.
The metrics worth pulling are win rates, conversion rates at each stage, average deal size, and sales-cycle length. And those numbers can vary quite a bit depending on how your team operates. In one benchmark analysis, teams applying RevOps best practices reported 87% higher win rates and 21% shorter sales cycles.
Now check whether your current pipeline backs up what the historical numbers suggest. Go through your open opportunities and look for anything that could move your forecast.
Start with whether enough new deals are entering the pipeline to support future sales, then look at where deals sit by stage and how long they’ve been sitting there. Watch for repeatedly pushed close dates and forecast categories that no longer match what's happening with the buyer.
If half your forecast depends on two opportunities, you want to know that before either one slips. And don’t assume a deal will close just because it reached a late stage. A deal can be in negotiation while the buyer is still working through approvals or hasn’t committed to a timeline. Your pipeline should tell you how realistic your revenue is, on top of how much of it is sitting in the CRM.
Learn how to spot gaps in your pipeline, diagnose revenue problems, and figure out what to fix with our RevOps 201 course 👇
This is where you separate the deals that look promising from the ones with a real shot at closing.
For every significant opportunity, ask these questions.
Base your answers on what the buyer has confirmed, not just what your sales team expects.
With your history and pipeline reviewed, you can start putting numbers together. Instead of leaning on one method, combine several so you get a fuller picture.
Historical forecasting sets your baseline from past performance and sales-cycle trends. Weighted pipeline forecasting values your open opportunities by how likely they are to close, and bottom-up forecasting uses deal reviews and forecast categories to judge what can realistically land. Funnel forecasting rounds it out by projecting future bookings from lead volume, conversion rates, win rates, and average deal size.
Calculate a few outcomes instead of betting on one number. A best case shows what happens if your larger opportunities close inside the period. A most likely case reflects what the buyer evidence actually supports, and a worst case shows what happens if important deals slip.
A single weighted forecast can mislead you. It accounts for the probability of a deal closing without showing how much your results hang on that one deal. So show people the number along with what could change it and by how much.
When the forecast period ends, go back and compare what you predicted with what happened. See how close you got to the final result, which deals closed as expected, and which ones slid into another period. Check whether certain forecast categories ran too optimistic, and whether your closing probabilities matched real outcomes.
One way to measure forecast accuracy is to compare your predicted sales against your actual results using this formula:
Look for patterns instead of treating each miss as a one-off. Forecasting gets more accurate when you learn from your previous calls.
No single sales forecasting technique is the most accurate in every situation. And accuracy is something most sales teams struggle with as only 7% of sales teams achieve forecast accuracy of 90% or more.
Combining historical performance with deal-level buyer evidence generally provides a more reliable forecast than relying on one method alone.
A reliable forecast needs three separate answers for each opportunity.
Look at whether the buyer has confirmed budget and whether you’re talking to someone with decision-making authority. There should also be a real business need behind the purchase. An enthusiastic buyer is a good sign, but enthusiasm doesn’t come with budget or sign-off, so judge the deal by what the buyer has done and give what they’ve said less weight.
Even when you’re confident a deal will close, the timing can be a different story. And with 57% of sales professionals saying the sales cycle is getting longer, predicting when a deal will close isn't getting any easier.
A buyer might be ready to move forward but still have legal reviews or budget approvals holding things up. That's why you need to confirm what's left before signature and whether those steps can realistically happen within your forecast period.
Closing a deal and earning revenue from it can happen at different times. Track the expected signature date, engagement start date, and when revenue will actually be earned. That way, your forecast reflects the revenue metric you're measuring instead of assuming everything happens when the contract is signed.
Good forecasting starts with good data, but a CRM full of numbers doesn’t mean you have what you need. And adding AI to the equation doesn't change that as 84% of data and analytics leaders say AI output is only as good as its underlying data.
You still have to know how your business has performed, what’s happening in your pipeline right now, and whether buyers are moving toward a purchase.
That comes down to three kinds of data:
Historical data gives your forecast a baseline. It shows how your sales process has performed over time and gives you somewhere to start when estimating what’s next. These are the metrics worth watching.
Metrics like average deal size and conversion rates are only useful if you can calculate them consistently. Our new BLUF software lets you build custom calculations using your CRM data, including metrics from different objects, so you can track the numbers that feed into your sales forecasts 👇
Historical data tells you what happened before, and your current pipeline tells you what’s happening now. Total pipeline value isn’t enough on its own. You need to know whether the pipeline can deliver the revenue you’re expecting.
Your CRM might show a deal in negotiation with a close date at the end of the month, but that doesn’t tell you whether the buyer can complete the purchase by then.
Your CRM is a good starting point but shouldn’t be your only source. Deal stages, close dates, and forecast categories are useful, but they don’t always tell the whole story. Call recordings show what buyers said about budgets, approvals, timelines, and next steps. Emails keep a record of buyer commitments and agreed dates, and account-team conversations add context about where the risks are.
A busy pipeline isn’t necessarily a healthy one. You can rack up plenty of meetings, emails, and follow-ups without the buyer getting any closer to deciding, so activity alone makes a poor measure of forecast confidence. What matters is whether the buyer is moving through the purchasing process.
MQLs, SQLs, and qualified opportunities all feed into your forecast, but each one tells you something different. MQLs and SQLs help you estimate future pipeline, while qualified opportunities give you a better read on near-term bookings.
The difference comes down to how far each prospect has moved through the buying process and how long it typically takes to turn that interest into revenue. You can use the funnel-based forecasting formula covered earlier to estimate future bookings from lead volume and conversion rates, but you'll also need to account for timing.
Marketing-qualified leads (MQLs) are leads Marketing has flagged as meeting its qualification criteria. They give you an early look at how much pipeline your marketing efforts could produce, but an MQL is still a long way from revenue. To estimate what they’ll contribute, look at how often MQLs become qualified opportunities and how long that usually takes. If your history shows 10% of MQLs eventually turn into opportunities, you can apply that rate to your current MQL volume to estimate the pipeline it might create.
Sales-qualified leads (SQLs) are leads Sales has decided meet its own criteria. They’ve made it further through qualification, so they give you another way to estimate future pipeline. What matters is how often SQLs become real opportunities. SQL volume tells you what’s coming and shouldn’t be counted as revenue already on its way.
Qualified opportunities are the most concrete evidence you have for near-term forecasting, since they’ve moved past lead qualification and into an actual sales process. You can look at each deal’s value, likelihood of closing, expected close date, and what the buyer still has to do. That makes them the right input when you’re forecasting the current month or quarter. Being qualified doesn’t guarantee a deal closes on schedule, though, so you still need to check buyer evidence and timing before it goes into your forecast.
If Marketing and Sales disagree on what makes a lead qualified, your forecasting numbers get shaky fast. Marketing might call a lead qualified based on engagement and fit, while Sales wants more proof of purchase intent before accepting it. Neither definition is automatically wrong, but both teams need to know exactly what each lifecycle stage means. Otherwise you could be applying historical conversion rates built on one definition to leads that are now being qualified under another, which makes your math look precise without making it accurate.
Double counting happens when the same potential revenue shows up in your forecast more than once. When a lead becomes an opportunity, its expected revenue should move over to the opportunity forecast and stop living on as a lead-based projection. You want your lead and pipeline forecasts working as one revenue model, with no stack of overlapping projections for the same deals.
Learn how to connect your leads, pipeline, and revenue data to make better decisions with our Revenue Performance Modeling course 👇
Even with good techniques and plenty of data, it’s surprisingly easy to get the numbers wrong. Here are seven common mistakes that throw off your numbers and what to do about each one.
Your target is what you want to hit, and your forecast is what the evidence says you’ll probably hit. They don’t have to match, and plenty of the time they won’t. The mistake comes when you tweak the forecast to look closer to the target instead of admitting the gap.
To avoid it: Build the forecast from history, pipeline data, and buyer evidence, then use any shortfall to figure out what needs to change.
Reaching the negotiation stage doesn’t guarantee a close. A deal can sit in negotiation while the buyer waits on budget approval. Pipeline stage tells you where the deal sits in your process, while forecast confidence tells you how likely it is to close in a specific period, and the two often disagree.
To avoid it: Judge confidence separately from stage by looking at buyer commitments, outstanding approvals, and realistic timelines.
Everyone loves an excited buyer, but your main contact might be completely sold and still have no authority over the budget or the signature.
To avoid it: Confirm who can approve the purchase and what approvals remain, and check whether the buyer has a realistic path to finishing them inside your forecast period.
Your buyer probably isn’t planning their purchasing process around your quota, so a close date a rep typed in only means something if it matches when the buyer can really decide.
To avoid it: Base close dates on timelines the buyer has confirmed and the steps left before signature.
A forecast isn’t something you calculate once and leave alone until the quarter ends. Default probabilities tied to pipeline stages are a starting point, but they shouldn’t replace ongoing judgment. Tools like HubSpot help teams stay on top of changing projections, refreshing AI sales projections five times per month (days 1, 7, 14, 21, and 28).
To avoid it: Revisit probabilities and categories whenever new information comes in. A buyer confirming budget might raise your confidence, while a surprise procurement delay might lower your confidence that the deal closes this quarter.
Bookings, recurring revenue, and recognized revenue measure different things, so combining them can make your forecast misleading even when the individual deal details are correct.
To avoid it: Define what you're forecasting before you start, track each metric separately, and make sure everyone knows which number they're looking at.
The review should help you understand what’s likely to happen, what’s putting the forecast at risk, and what the team can do about it. Try asking what the biggest remaining risk is to each deal and what evidence says the buyer can decide within the forecast period. Then ask what one action could reduce that risk this week and who owns it.
To avoid it: Make every review about finding uncertainty and acting on it, so you leave with a clearer forecast and a concrete plan for the biggest risks.
Here's what matters most:
Sales forecasting is the process of estimating how much revenue a business will generate over a specific period, typically a month or quarter. It uses historical sales performance, current pipeline data, and buyer insights to predict which deals are likely to close and when. Accurate forecasting helps businesses plan budgets, allocate resources, and identify potential revenue shortfalls.
The four main sales forecasting methods are historical forecasting, weighted pipeline forecasting, bottom-up forecasting, and funnel-based forecasting. Historical forecasting uses past results, weighted pipeline forecasting applies closing probabilities to deal values, bottom-up forecasting evaluates individual opportunities, and funnel forecasting projects future bookings using lead volume and conversion rates. Combining methods generally provides a more reliable picture than relying on one alone.
The formula depends on the forecasting method. For weighted pipeline forecasting, multiply each opportunity's value by its probability of closing, then add the results. For funnel-based forecasting, multiply lead volume by the conversion-to-opportunity rate, opportunity win rate, and average deal value. Both methods require accurate data and adjustments for when deals are expected to close.
Accurate sales forecasting requires three types of data: historical sales performance, current pipeline health, and buyer evidence. Important metrics include win rates, average deal size, sales-cycle length, deal stage aging, and expected close dates. Teams should also verify buyer budgets, decision-making authority, outstanding approvals, and purchasing timelines rather than relying solely on CRM deal stages.
Improve forecasting accuracy by combining historical data with deal-level reviews, regularly updating closing probabilities, and validating timelines against buyer commitments. Build multiple forecast scenarios to understand the impact of delayed or lost deals, and distinguish between bookings, recurring revenue, and recognized revenue. Finally, compare forecasts against actual results to identify patterns and refine future predictions.