I sit in on a lot of forecast calls. Most of them are some of the least useful hours I’ve ever spent in a sales organization.
The pattern is almost universal. The VP of Sales opens with this quarter’s number. The directors give status on their regions — usually within a percentage point of where they were last week. Each rep walks through their top three deals and gives an opinion on whether they’ll close. The CRO writes down a number. The number is, more often than not, wrong by ten to twenty percent.
This is not a forecasting process. It’s status theater dressed up as governance.
Gartner research suggests fewer than half of sales leaders have high confidence in their forecasts, and that the median forecast accuracy across surveyed organizations sits between 70 and 79%. Other industry analyses peg the slippage rate even higher, with around 60% of deals tagged as “commit” missing the quarter they were originally forecasted in. For a company running on capital efficiency, that’s not a forecasting problem — that’s a planning problem with cascading effects on hiring, capital allocation, and board credibility.
Why the Standard Forecast Call Doesn’t Work
There are three structural reasons most weekly forecast calls produce bad numbers.
The rep’s frame dominates. A forecast call that asks “Is this deal going to close?” is asking the most biased person in the conversation. The rep believes in their deal — that’s a feature, not a bug, of being a good salesperson. But it makes their forecast input nearly useless without an external check.
The conversation centers on the deal, not the evidence. “It’s looking good” and “the champion is engaged” are not data points. They’re vibes. When forecast calls trade in vibes, the number becomes whatever the most confident voice in the room says it is.
Nothing changes between weeks. Most forecast calls end with the same deals in the same categories with the same probabilities. A forecast process that produces no new decisions is a meeting, not a forecast.
“If your forecast call ends with the same numbers it started with, it isn’t a forecast call — it’s a meeting.”
— Dalton Ezri
The Three-Tier Cadence That Works
The forecast process I install with clients has three distinct conversations, on three different cadences, with three different purposes. Confusing them is the most common mistake.
Tier 1: Weekly deal review (operational). This is where individual deals get worked. Manager and rep, thirty minutes, four to six deals max. The conversation is structured around evidence: what do we know, what are we assuming, and what are the named actions to close the gap. This is where coaching happens. It is not a forecast conversation.
Tier 2: Bi-weekly forecast call (analytical). Director-level and above. The conversation is about the forecast as a whole — which segments are over-pacing, which are under, where the assumption gaps sit, and what would need to be true for the number to come in. Individual deals only come up as examples or evidence. This is where the number gets sharpened.
Tier 3: Monthly executive forecast (strategic). CRO with the CEO and CFO. Three numbers: the call, the worst case, and what we’re doing to close the gap between them. No deal-by-deal narrative. The conversation is about resource allocation and risk.
When I install this cadence, the most common pushback I hear is that it’s “too many meetings.” It isn’t. Most organizations are already doing all three conversations — they’re just mashing them into a single weekly call where none of them get done well.
The Stage-Gate Evidence Requirement
The single biggest lever for forecast accuracy isn’t the cadence — it’s what’s required to advance a deal between stages. Stage gates that require named evidence (a champion confirmed in writing, an economic buyer meeting taken, a procurement timeline confirmed) produce dramatically different forecasts than stage gates that rely on rep judgment.
I’ll often install a “two confirmed criteria minimum” rule for any deal in the commit category. That is: the deal can’t be called commit unless we have at least two named, verifiable pieces of evidence — not assumptions, not feelings — supporting that it closes this quarter.
The first month after installing that rule is usually painful. Commit pipeline shrinks by 30 to 50%. Leaders panic. But the real commit pipeline was always smaller than the reported one — the rule just exposes the gap. From that point on, you’re forecasting reality instead of optimism, and accuracy improves quickly.
As I often tell new CROs: the goal of a forecast process isn’t to inflate confidence. It’s to produce a number you and your CFO can both stake your credibility on, week after week.
The Named-Action Close
Every forecast conversation, at every tier, should end the same way: with named actions, named owners, and named deadlines.
Not “we’ll keep an eye on it.” Not “let’s revisit next week.” Specifically: who is doing what, by when, that will materially change the probability of this deal — or this segment — closing as forecast.
A forecast call without named actions is a forecast call you didn’t need to have. The artifact of a good forecast process isn’t the spreadsheet — it’s the list of decisions and actions that came out of the conversation.
“A forecast process that produces no new decisions is a forecast process that produces no new accuracy. The two are the same problem.”
— Dalton Ezri
The Trajectory
For most teams I work with, accuracy moves from somewhere in the 60–70% range to the 80–85% range within two quarters. Some teams get to 90%+ in the second year. None of this happens overnight — the cultural shift from “rep opinion” to “evidence-backed assumption” takes time, and there is always a quarter where the forecast comes in lower than it would have under the old process. That’s not a failure. It’s the moment the forecast finally became real.
Companies that install structured forecast coaching see forecast accuracy improvements of up to 15 percentage points, according to Gartner. That’s not a marginal gain. That’s the difference between a CRO who can credibly run a planning cycle and one who can’t.
If your forecast call has felt the same every week for the last year — same deals, same opinions, same near-miss outcomes — the call isn’t broken. The cadence underneath it is.
Dalton Ezri works with sales leadership teams to redesign forecast processes that actually predict revenue — from cadence design to stage-gate evidence requirements to the executive conversation that holds the whole thing together. If your forecast accuracy isn’t where it needs to be, the answer is rarely a new CRM and almost always a redesigned process.
