On Sunday I argued that revenue forecasts break upstream, in the middle, and across functions. This is the part I’m less certain about: what I’d actually do about it. Argue with me.
Sunday’s edition made a case I feel reasonably confident about. Only 45 percent of sales leaders have high confidence in their own forecasts, according to Gartner. Somewhere between 40 and 60 percent of deals end in no decision rather than a competitive loss, according to Matt Dixon and Ted McKenna‘s analysis of 2.5 million sales conversations. And 95 percent of winning vendors were already on the buyer’s day one shortlist, according to Kerry Cunningham‘s 2025 Buyers Process at 6sense, which means most of the quarter was decided before anyone in sales knew a deal existed. I’ve tried to put these all together in my mind to make sense of what is going on and ultimately how to improve our forecasting accuracy and reliability.
Diagnosis is the easy part. Here’s the harder question: what should we actually do about it?
What follows is a model I’ve been working on. I’ve used pieces of it. I’ve not run the whole thing across a portfolio, and I’d genuinely like to know where it breaks.
The Model
It’s a sibling to my Customer Cohort Triangulation model that analyzes win/loss, GRR, and NRR percentages to align short- and long-term growth, and the underlying idea is the same. Don’t trust one source. Look at the same thing from three independent angles and pay attention to where they disagree.
Three views of the same quarter. Three different owners.
The seller view. Bottom-up rep commit and best case, rolled up. The CRO’s number. It answers: what do the people actually in these deals believe?
The math view. Historical conversion by stage and segment applied to current pipeline, adjusted for aging and slip history. RevOps owns it. It answers: what has pipeline that looks like this actually done before?
The brand and demand view. Account engagement, intent signals, ICP fit of what entered, earned versus bought sourcing, cohort quality by segment. The CMO’s number. It answers: is real buying happening upstream, and is what’s entering the funnel any good?
When the three converge, I’d argue you will have a more reliable revenue forecast. When they diverge, the direction of the divergence may tell you what’s actually broken.
Reading the Divergences
Seller view above the math view suggests inspection is failing. Deals sitting in stages they haven’t earned, or optimism reported as commitment. Probably a CRO fix, and probably a fast one.
Math view above the demand view suggests the pipeline is aging with nothing refilling behind it. This quarter might hold. The next two probably won’t. That’s a brand reputation problem that marketing needs to drive cross-functionally for investment to seed the right demand early, it takes two or more quarters to correct, and that’s exactly why you’d want to see it early.
Demand view above the seller view suggests real demand exists and isn’t being converted or worked. Capacity, enablement, qualification, routing, coverage, and execution. That’s an alignment problem rather than a function problem, and it likely needs the CEO in the room.
What appeals to me about this, and I’ll admit it may be the thing I like rather than the thing that’s true, is that you can’t triangulate without all three functions at the table. Forecast credibility stops being something the CRO owns and reports, and becomes something the go-to-market team produces together.
Who I’d Put in Charge of What
Shared ownership usually means nobody owns it, so let me be specific.
RevOps owns definitions and math. One coverage formula, published, agreed by finance, sales, and marketing. Stage exit criteria that mean something. Aging and slip tracking. And conversion rates refreshed regularly rather than set once, because buyer behavior has shifted enough in the last two years that stage definitions written in 2023 are probably wrong now.
The CRO owns the seller view and inspection. Deal-level rigor, honest categorization, and the willingness to close out deals that aren’t real. Removing pipeline is the hardest part of that job, and the CROs I’ve seen do it well have the most credible forecasts.
The CMO owns the brand reputation and demand view and everything upstream. What enters the funnel and whether it’s any good. ICP fit at the point of entry, not at the point of loss. Earned versus bought sourcing. Brand strength, because that’s at-bats. And forecasting marketing’s own number and hitting it, because I don’t think you earn standing in the pipeline conversation until you’ve earned it on your own.
The CEO owns making sure those three people are in the same room weekly. I’ve never seen it happen without that.
The Rhythm
Three cadences, three different questions.
Weekly, a four-week view. Deal-level inspection. What’s closing, what’s at risk, what moved and why. The CRO’s meeting.
Monthly, an eight to twelve week view. Cohort-level rather than deal-level. Is pipeline aging? Is what entered this month better or worse than last month? What does the math view say against the seller view? This is where triangulation actually happens, and all three functions need to be there.
Quarterly, a two to three quarter view. Coverage build, demand quality trend, whether the upstream drivers are moving. This is the board conversation, and it’s the one I see run properly least often.
Rebuilding Credibility in One Quarter
If credibility is already lost, here’s what I’d try, held loosely.
Publish the definitions first. One coverage formula, one set of stage exit criteria, one definition of qualified, signed off by all three functions. My experience is that this alone solves more than it should.
Then clean the pipeline. Close out everything stale, everything that’s slipped twice, everything failing the new qualification bar. Coverage will drop and it will look bad. I’d tell the board you’re doing it and why before you do it, because voluntarily shrinking your own pipeline number is the opposite of what people expect and it seems to buy real credibility.
Then forecast conservatively and hit it. Once, then twice. I don’t know a clever version of this. Credibility looks to me like a track record, and it accumulates one quarter at a time.
The Bottom Line
The forecast isn’t broken at the forecast. It breaks upstream, in the middle, and across functions, which is why no single leader can fix it alone.
The model I’d propose is triangulation. Seller view, math view, brand and demand view. Where they disagree is the diagnosis, and you can’t run it without all three functions in the room.
But I’ll say plainly that I’m still working this out. I’ve watched forecasting break in enough companies to feel confident about what’s wrong. I’m less confident I’ve landed on the fix.
So tell me where this falls apart. That’s a genuine ask, and I’ll write about what I learn.
Outcomes over activities. Every time.
Sources
Gartner, State of Sales Operations Survey, on forecast confidence and accuracy.
Kerry Cunningham and 6sense, 2025 B2B Buyer Experience Report, approximately 4,000 buyer responses across North America, EMEA, and APAC.
Matthew Dixon and Ted McKenna, The JOLT Effect, based on analysis of more than 2.5 million recorded sales conversations using the Tethr platform.
