How accurately can a company forecast its sales revenue today?

Larry Shurtz · May 10, 2024

Inability to forecast accurately is a symptom of deeper problems: not understanding the business, insufficient diligence on customers and opportunities, too few opportunities to balance risk, and asking the wrong questions of the wrong people

Accurate forecasting is downstream of real deal diligence and pipeline breadth, so misses reveal those gaps

0:00 20Sales: How to Build Vertical Sales Teams, Why No Customer Success is BS and Everyone Needs it, How to Hire, Train and Retain the First Reps and Lessons Scaling to $2.1BN Revenue and 1,300 People with Larry Schurtz, CRO @ Genesys

Larry Shurtz · May 10, 2024

Forecast accuracy is a core part of a sales leader's brand, and chronic inaccuracy signals you don't understand your business, aren't doing enough diligence, lack enough pipeline to offset risk, or are asking the wrong questions of the wrong people

The company's investment and strategic priorities depend on the revenue leaders owning the number, and the roll-up only works if every layer of directs down to the AEs does what they say they will do

Scope: it is genuinely hard to do well — he has missed months and back-to-back quarters; specific to the enterprise world he has chosen to play in

42:47 20Sales: How to Build Vertical Sales Teams, Why No Customer Success is BS and Everyone Needs it, How to Hire, Train and Retain the First Reps and Lessons Scaling to $2.1BN Revenue and 1,300 People with Larry Schurtz, CRO @ Genesys

Dave Kellogg · Jan 31, 2024

Reps should submit their own weekly forecast for the quarter and managers should never tell a rep what their forecast is

A manager dictating the number destroys the independence that makes triangulation work; you want the rep's number to be what the rep believes and the manager's number to be what they believe their reps will do, and weekly submissions let you watch the number move

27:09 20VC: The Metrics That Matter in SaaS Today; Why CaC Payback is Flawed & CAC Ratio is Better, Why You Need to Hire Three Sales Reps at a Time, How to Forecast in 2024 & Biggest Mistakes Made Forecasting & How to Make Customer Success Sell More with Dave K

Dave Kellogg · Jan 31, 2024

Forecasts should never be stretch numbers — a figure a manager has pressured a rep into committing to is useless for predicting the business

A forecast is purely a prediction of what will be sold; when managers use it as a club to extract a bigger commitment in a moment of weakness, you lose the honest number you actually needed

29:18 20VC: The Metrics That Matter in SaaS Today; Why CaC Payback is Flawed & CAC Ratio is Better, Why You Need to Hire Three Sales Reps at a Time, How to Forecast in 2024 & Biggest Mistakes Made Forecasting & How to Make Customer Success Sell More with Dave K

Carles Reina · Apr 11, 2026

Accurate sales forecasting is essentially impossible in this environment

If you run the business as a portfolio of experiments, you only need one experiment to land to add a million or two hundred million in revenue, which makes outcomes unpredictable

24:58 Experiment portfolio makes outcomes unforecastable

Harry Stebbings · Apr 11, 2026

Traditional sales leaders' claim that they can forecast to within 3-5% now looks impossible

Twenty-seven episodes of sales leaders said they could hit that accuracy, but the current environment contradicts it

25:17 Traditional tight forecast accuracy no longer holds

Sean Murray · Jan 24, 2024

In 2024 you should forecast with AI/machine-learning tooling rather than human prediction

The technology puts real science into 90-day-out output — on his new logos side it has been within 5% accurate — whereas he has spent years trying to predict the future manually and still can't get it right

58:02 Ai ml forecasting tools outperform human prediction in 2024

Frank Fillmann · Mar 22, 2023

Forecasting under uncertainty should combine three inputs: a historical math model of close rates to pipeline, AI-based trend detection at scale, and a bottoms-up qualitative deal-by-deal review that is trued up against the other two

Technology can pattern-match across thousands of reps in ways a small team cannot, and comparing the math and AI outputs against a human deal review tests whether you believe the numbers

33:51 Combine historical model ai trend detection and qualitative review to forecast

Chad Peets · May 23, 2026

Forecasting must stay grounded in a data-driven productivity model — new ACV per rep, hiring, attrition and ramp rates — with outlier deals layered on top, rather than being replaced by feel

That model historically predicts year-end with about 90% certainty; companies raising on 'we just feel like we can do 300' can't back the number up

32:46 Capacity and productivity math still forecasts accurately

Stevie Case · Jan 25, 2023

The hardest part of the CRO job today is having to commit to a year-out plan when conditions change so rapidly

The pace of change is rapid but planning horizons are annual, and she wants enough foresight to avoid changing course abruptly

43:25 Annual planning horizons conflict with rapidly changing conditions

Jason Lemkin · May 6, 2022

Long enterprise sales cycles stop mattering once you're north of $20-30M in revenue because you have enough deals in flight to be predictable

With a hundred real deals in flight, outcomes start to happen predictably regardless of cycle length; almost no startup founders have that many enterprise deals in flight

17:43 Sufficient deal volume in flight makes outcomes statistically predictable regardless of cycle length

Ben Fiechtner · Aug 2, 2024

For forecast accuracy, process matters as much as or more than tooling — specifically a thirteen-week revenue cadence broken into segments where everyone reviews the same data

Shared data means frontline managers all ask the same questions, AEs know what to expect, and senior leaders can manage by exception to locate whether the hole is top of funnel, conversion, execution or churn; most forecast calls are just reps saying 'this deal's still good'

14:58 Structured weekly cadence process matters as much as tooling for forecast accuracy

Ben Fiechtner · Aug 2, 2024

Best practice is to require leaders' penciled forecast number to be backed by a commit deal walk-up that ties out — a million-dollar commit needs a million dollars of commit deals

It forces leaders not just to call a number but to call their shot on which specific deals are furthest along

17:10 Commit forecast must tie out to named commit deals

Ben Fiechtner · Aug 2, 2024

The core value of a forecasting platform for a CRO is signal-to-noise: telling you exactly where the gap in the business is so you know where to spend your limited time.

A CRO — like a founder — has a gun to their head every day and not enough hours, so the tool's job is to point at the five deals that won't close.

23:25 Forecasting tool value is signal to noise locating the gap

Dave Kellogg · Jan 31, 2024

Churn forecasting is substantially easier than new sales forecasting

For churn you have product usage data showing whether usage is rising or falling, whether data is being dumped out, whether a competitor is being spun up, plus direct relationships where you can simply ask the customer; new sales forecasting means competing in a bake-off where sophisticated buyers deliberately withhold information and a committee of constituents may not itself know who is winning

25:32 Churn forecasting is substantially easier than new sales forecasting

Dave Kellogg · Jan 31, 2024

A good forecast is a gently upsloping weekly curve that lands on the number you actually sell; forecast quality is about accuracy, not quota attainment

Landing where you said means you are not misleading your boss or, in aggregate, the company; wildly optimistic forecasts that collapse make the business impossible to run, and extreme lowballing is also hard to run against

28:25 Forecast quality is accuracy to actual not quota attainment

Dave Kellogg · Jan 31, 2024

A forecast call should be all numbers, and a forecast call that is all storytelling is a sign to fire the sales manager

The purpose of the call is to produce the forecast — the question asked is how much you're going to sell this quarter, not a narrative of who talked to whom

34:12 Forecast calls should be numbers only storytelling signals a bad manager

Kyle Norton · May 30, 2025

A revenue leader must model every input sell-by-sell and manage six to twelve months ahead with deliberate channel redundancy, or they will be blindsided

Pipeline shortfalls are unrecoverable when discovered in-quarter — with a 120-day cycle the mistake was made six months earlier — and one channel always falters, which is why first-time VPs of sales get cooked

30:57 Revenue leaders must model every channel input months ahead with redundancy to avoid blindside

Patrick Forquer · May 11, 2026

In a business growing this fast, boards should be guided on a rolling six-month basis rather than full-year projections

Projections are very hard when the company blew through 25 in Q3 and ended the year at 70; only the near-term pipeline data is reliable enough to forecast against

48:43 Hypergrowth limits guidance to a rolling six months

Patrick Forquer · May 11, 2026

Accurate forecasting requires running a 'bet your life' rep-to-manager-to-VP commit roll-up and comparing it against a separately weighted, stage-based math forecast

When the two independent methods converge you can trust the number; last quarter the regional roll-up and the weighted forecast matched and both were accurate

50:07 Two independent methods converging validates the forecast