The Two Types of Sales Forecasts Every CRO Needs (and When to Use Each)

If you've ever sat in a board meeting where your CFO asks "Can we hit the number?" while your board member asks "Should we open that new region?"—you've witnessed the fundamental forecasting problem in B2B revenue.

They sound like similar questions. They're not.

One demands certainty about the next 90 days. The other requires expected value thinking across 12-18 months. Yet most revenue leaders try to answer both with the same weighted pipeline number, and that's where the trouble starts.

Here's what typically happens: You pull up your CRM, filter for all opportunities closing this quarter, and present the weighted pipeline value—let's say $8M weighted down from $12M total pipeline. Your CFO nods. Your board notes it. Everyone moves on.

But here's the problem: that $8M number is simultaneously too optimistic for quarterly planning and too conservative for annual capacity decisions.

For quarterly execution, you're including $3M worth of 60% probability deals that probably won't close this quarter (and might never close). For annual planning, you're ignoring the upside case where multiple deals come in larger than expected or new pipeline materializes.

You're using portfolio math where you need binary thinking, and binary thinking where you need portfolio math.

Portfolio thinking means treating your pipeline like an investment portfolio: every deal contributes its expected value (deal size × probability), and across many deals, the math normalizes.

If you have 200 deals in your annual pipeline:

• 50 deals at $100K with 30% probability = $1.5M expected value
• 75 deals at $200K with 50% probability = $7.5M expected value
• 50 deals at $300K with 60% probability = $9.0M expected value
• 25 deals at $500K with 70% probability = $8.75M expected value

Total expected value: $26.75M

This is legitimate math. Across 200 deals, the probabilities will largely play out as expected. Some 30% deals will close. Some 70% deals will slip. But the aggregate expected value will be reasonably accurate.

When to use portfolio thinking:

• Annual and multi-quarter revenue planning
• Territory design and quota setting
• Headcount and capacity modeling
• New market investment decisions
• Long-term financial projections

The key is sample size and time horizon. With enough deals over enough time, probability distributions normalize and expected value becomes predictive.

Binary thinking means only counting deals that have crossed a high-confidence threshold—typically 80% or higher probability of closing in the specified time window.

For that same pipeline, if only 40 deals meet your 80%+ criteria:

• 15 deals at $200K = $3.0M
• 18 deals at $300K = $5.4M
• 7 deals at $500K = $3.5M

Total high-confidence forecast: $11.9M

Notice this is less than half your expected value number. That's not sandbagging—that's probability math over a short time window.

When to use binary thinking:

• Monthly and quarterly forecasting
• Board commit numbers
• Compensation triggers
• Near-term cash planning
• Resource allocation for current quarter

Why? Because in a 90-day window, you don't have enough sample size for probabilities to normalize. Ten deals at 70% probability doesn't mean seven will close—it means you have massive variance. You might close three, you might close nine.

This is where most forecasts break down. Let's illustrate with a simple example:

Imagine you have 10 coins that each have a 70% chance of landing heads. If you flip all 10, you'd expect 7 heads on average. But run this experiment 100 times, and you'll see:

• 7 heads (the "expected" outcome): happens about 27% of the time
• 6 or 8 heads: each happen about 23% of the time
• 5 or 9 heads: each happen about 10% of the time
• 4 or fewer, 10 heads: combined happen about 7% of the time

So the "expected outcome" only happens about a quarter of the time. The actual result has a massive range.

Now apply this to your quarterly forecast. If you have 15 deals at 60-70% probability, your weighted forecast might say $3M. But the actual close rate could easily range from $1.8M to $4.2M—and both would be perfectly consistent with those probabilities.

This is why treating quarterly forecasts as probabilistic is dangerous. The variance is too high.

The biggest mistake in sales planning is treating deal stages as probability proxies. "It's in Negotiation, so it's 80%." No—stage progression and close probability are related but not identical.

A deal should only hit 80%+ confidence when you have objective validation:

For enterprise deals:

• Signed MSA or paper is out for signature
• Budget explicitly confirmed and allocated
• Technical validation complete (POC passed, security review done)
• Economic buyer personally engaged and advocating
• Procurement process mapped with no unknown gates
• Legal redlines are minor and agreed upon

For mid-market deals:

• Champion has confirmed budget and timing in writing
• You've presented to the economic buyer
• No competing vendors remain in active evaluation
• Contractual terms agreed (not just discussed)
• Implementation timeline aligned with their planning

Notice these are binary gates, not probability estimates. A deal either has executive sponsorship or it doesn't. Budget is either confirmed or it's "probably there."

When you pressure-test deals this way, far fewer hit 80% than you think. And that's good—it's better to know now than at month-end.

Countless teams fall into this pattern:

Week 1: "We have $5M weighted pipeline for the quarter" Week 6: "Still tracking $5M weighted, we added some new deals to replace slippage"
Week 10: "At $4.8M weighted, we'll close $4.5M" Week 13: Closed $3.2M

What happened? They were treating a small sample of deals as if probability would normalize in 90 days. It doesn't.

The worst version of this is recalculating probabilities upward as the quarter progresses to maintain the illusion of pipeline coverage. "Well, if these deals don't close, we'll miss the quarter badly, so they must be higher probability than we thought." This is magical thinking.

The real problem: When you build quarterly forecasts on 50-60% deals, you're not just accepting variance—you're hiding from reality. You're counting deals that probably won't close and pretending the math will save you.

But binary thinking has its own failure mode. Revenue leaders have been known to use 80%+ criteria for annual planning and systematically underinvest in growth.

Example: A company wants to open EMEA. They build a pipeline of 30 opportunities over six months—all at 40-60% probability because it's a new market with unproven messaging and case studies. Using binary thinking, none of these count, so the region "isn't working" and doesn't get more headcount.

But portfolio thinking says: 30 deals × $250K average × 50% probability = $3.75M expected value. That might justify two more AEs and an SE, which could push probabilities higher and generate more pipeline.

The innovation penalty: New products, new markets, and new segments never have 80% deals early. If you only count high-confidence opportunities in long-term planning, you'll never invest in anything new.

Similarly, annual capacity models built only on "sure things" lead to chronic under-hiring. You miss your hiring window, then scramble to add headcount mid-year when "certain" pipeline materializes into closed deals and you lack capacity to work new opportunities.

The solution isn't choosing between portfolio and binary thinking. It's running both simultaneously and being explicit about which one you're using for which decision.

Your weekly pipeline review:

• Portfolio view: "We have $18M in expected value across 180 opportunities for the next two quarters"
• Binary view: "We have $4.2M in high-confidence commit for this quarter across 22 deals"

These numbers tell you different things:

• The $18M tells you if you're generating enough pipeline for sustained growth
• The $4.2M tells you if you'll hit this quarter's number

Both are true. Both are useful. They answer different questions.

To implement portfolio thinking rigorously, you need historically calibrated probabilities, not stage-based guesses.

Step 1: Calculate actual conversion rates Pull 12-18 months of historical data and calculate conversion rates by:

• Sales stage
• Deal size band
• Segment (Enterprise, Mid-Market, SMB)
• Rep tenure (new hires vs. veterans)
• Competitive situation
• Time in stage

You'll find that "Negotiation" might be 65% for enterprise deals with Darsha (your top rep), 45% for mid-market deals with new reps, and 80% for renewals. This granularity matters.

Step 2: Apply calibrated probabilities Take your current pipeline and apply the historical conversion rates. Don't let reps set their own probabilities—use your data model. A deal in Negotiation gets the conversion rate for deals in that stage with that profile, not whatever the rep thinks.

Step 3: Add confidence intervals Run Monte Carlo simulations to understand variance. You might have $12M expected value, but there's:

• 70% confidence you'll land between $10M-$14M
• 90% confidence you'll land between $9M-$15M
• 95% confidence you'll land between $8M-$16M

This gives your CFO and board a realistic picture: "We expect $12M, but there's normal variance of $2M in either direction."

Step 4: Update based on deal velocity Adjust probabilities based on engagement signals:

• Days since last meaningful customer interaction
• Slippage from original close date
• Stakeholder engagement scores
• Competitive displacement risk

A deal that's been in Negotiation for 60 days with no executive engagement might need its probability cut in half, regardless of stage.

The binary forecast requires human judgment combined with objective criteria.

Step 1: Define the 80% threshold Create a scorecard of required elements. For enterprise deals, it might include:

• [ ]  Paper out or MSA signed
• [ ]  Budget confirmed by economic buyer in writing
• [ ]  Technical validation complete
• [ ]  Executive sponsor engaged in last 14 days
• [ ]  Legal redlines resolved or minor
• [ ]  No competing vendors in active evaluation
• [ ]  Close date within 30 days

If a deal doesn't check every box, it doesn't qualify for commit—even if it's in "Closed Won" stage.

Step 2: Implement rigorous deal reviews Weekly reviews where managers pressure-test every deal in the commit forecast:

• "When did you last speak to the economic buyer?"
• "What happens if their Q1 budget gets cut by 20%?"
• "Which executive is putting their credibility on the line for this deal?"
• "What could cause this to slip to next quarter?"

The goal is not to be pessimistic—it's to be accurate. Reps should welcome this rigor because it helps them see risks they can mitigate.

Step 3: Track accuracy religiously Every week, compare:

• What was in high-confidence forecast
• What actually closed
• What slipped and why

Your binary forecast should close at 85-95% accuracy. If you're at 70%, your threshold isn't high enough. If you're at 99%, you're sandbagging.

The hardest part is translating between these models for different stakeholders.

For your board: "Our expected value for the year is $48M, which gives us confidence in the $45M plan. For Q1 specifically, we have $11M in high-confidence commit against a $12M quota. The gap is typical for this point in the quarter—we usually close another $1.5-2M in the final two weeks from deals that accelerate."

For your CFO: "For the quarter, I'm committing to $11M. We have $19M in total pipeline weighted to $13.5M expected value, but I'm only counting deals that meet our high-confidence criteria. Here's the list of 27 deals and their status."

For your sales team:
"Your territory has $3.2M expected value for the year based on current pipeline and historical conversion rates. That suggests you can achieve your $2.8M quota. For this quarter specifically, you have four deals totaling $850K that meet our commit criteria. Let's talk about what it takes to get two more deals into that category."

Notice how the language changes based on the question being asked.

Pitfall 1: Averaging the methods Some teams try to split the difference: "Well, portfolio says $8M and binary says $4M, so let's forecast $6M." This gives you the worst of both worlds—you're still overcommitting on probability deals while underinvesting in long-term growth.

Pitfall 2: Letting reps choose
"You can either forecast based on weighted pipeline or commit-level deals, whatever makes sense for your territory." No. The method must be consistent across the team, or you have no idea what your aggregate forecast means.

Pitfall 3: Switching mid-quarter When the quarter looks bad, teams sometimes switch from binary to portfolio thinking: "Well, if we count all the 60% deals, we could still make it." This is desperation, not forecasting.

Pitfall 4: Using portfolio math for comp If you pay accelerators based on "expected value delivered" rather than actual bookings, you'll pay out on deals that never closed. Compensation should always use binary thinking—pay on actual results.

Getting your team to operate in both modes requires consistent practice and cultural reinforcement.

Weekly practice: Have every rep articulate both numbers for their territory:

• "I have $1.8M expected value across 14 deals for the quarter"
• "I have $650K in high-confidence commit across three deals"

Then ask: "What needs to happen to get deal #4 into the commit category?" This trains reps to think about both probability and confidence simultaneously.

Calibration exercises: Track rep accuracy on both:

• Were their probability estimates correct in aggregate? (portfolio accuracy)
• Did their high-confidence deals close at 85%+? (binary accuracy)

Reps who consistently overestimate probabilities need coaching on qualification. Reps who consistently sandbag binary forecasts might be playing comp games.

Manager coaching points:

• "This deal is in your commit forecast, but they haven't confirmed budget. That's inconsistent."
• "You have $2M expected value but only $400K in commit. What's blocking the other deals from accelerating?"
• "Your commit forecast has been 95% accurate for four quarters. You might have room to include one more deal."

Cultural shift: Celebrate accurate forecasting, not just beating the number. A rep who forecasts $800K and delivers $820K is more valuable than one who forecasts $1.2M and delivers $900K. Accuracy enables planning.

Mature revenue organizations operate both systems in parallel without confusion. They know that annual planning requires thinking in expected values and confidence intervals. They know that quarterly execution requires high-confidence binary forecasts.

This sophistication shows up in board meetings where the CRO can say: "We're tracking to plan for the year based on portfolio math, but this quarter will be light because three enterprise deals pushed to Q2. Here's why I have high confidence they'll close next quarter."

It shows up in hiring decisions where the VP of Sales can say: "The expected value in our new segment justifies two more reps, even though we don't have any 80% deals yet. At their typical ramp time, they'll have territory-level pipeline by Q3."

It shows up in forecast accuracy where the revenue team consistently delivers within 5% of their quarterly commit while accurately predicting annual outcomes within 10%.

The competitive advantage is real: better capital allocation, more reliable quarters, appropriate investment in growth initiatives, and credibility with your board and CFO.

Start by auditing your current approach:

1. What method are you using for quarterly forecasts? If you're counting 50-60% deals, you're using portfolio thinking where you need binary.
2. What method are you using for annual planning? If you're only counting 80%+ deals, you're using binary thinking where you need portfolio.
3. Do you have historically calibrated conversion rates? If not, you're guessing at probabilities rather than using data.
4. Can you articulate your high-confidence criteria? If "it depends on the deal," you don't have binary rigor.
5. How accurate have your forecasts been? If quarterly accuracy is below 85%, you need stricter commit criteria. If annual accuracy is off by more than 15%, your portfolio model needs work.

The best CROs know which language to speak in which conversation. They know when to embrace the uncertainty of probability distributions and when to demand the rigor of binary gates. They run both systems simultaneously because revenue forecasting isn't one problem—it's two fundamentally different problems that require two fundamentally different approaches.

The organizations that master this dual-mode thinking make better decisions, hit their numbers more consistently, and allocate capital more effectively than their competitors. That's the difference between a revenue leader who provides a number and a CRO who provides clarity.