Recapping our LinkedIn Live with Doug May, a 30-year go-to-market leader, and the Airspeed team, on why forecasts break down inside revenue organizations and what it actually takes to fix them. Missed it? Watch the full recording →
It’s the final weeks of the quarter. Your Salesforce commit is locked. The forecast is ready. The CRO feels good, the VPs feel good, the managers have inspected every deal, and the number has already traveled up the chain, from the CRO to the CEO to the board.
Then four deals slip.
Nobody lied. That’s the uncomfortable part. The rep didn’t set out to deceive anyone. But somewhere in the last two weeks, the economic buyer went quiet, a critical meeting got pushed, legal red lines were never opened, the security review never started. The signals were there. They just weren’t in the room when the number got called.
We built an entire webinar around this moment because it repeats itself, quarter after quarter, at companies that should know better. Airspeed’s VP of Sales Thang Nguyen sat down with Doug May, who spent three decades in go-to-market, sat in the CRO seat, and advised the CROs at Splunk, Databricks, Datadog, and Harness, plus Airspeed product engineer Alex Painter, to take the problem apart. Here’s what came out of it.
Forecasting isn’t a scorecard. It’s an investment decision.
Early in his career, Doug thought forecasting was a way to hold the sales team accountable: make sure people do what they say they’ll do. The lesson that reframed everything came from a CFO he worked with at Splunk: the company makes investment decisions based on the forecast.
Forecast conservatively, and the company won’t lean in. It won’t hire the developers, the legal reviewers, the sales engineers the business needs to grow. Forecast too high and miss, and the damage runs the other way. Doug has watched companies between $50M and $400M in revenue freeze hiring outright when the number didn’t land.
Either direction distorts the growth, the momentum, and ultimately the valuation of the company. And for the person holding the number, the stakes are personal: you can be a great recruiter and run great process, but if you can’t forecast, you won’t sit in the CRO seat for long.
The tools became glorified spreadsheets
Forecasting tools have evolved, Doug argued, mostly into calculators. They do the math so a second- or third-line leader doesn’t have to roll up 80 reps by hand. What they don’t do is add intelligence.
So the cadence takes over: forecast call after forecast call, down the chain and back up. And what the tool ultimately reflects is opinion layered on opinion layered on opinion. Thang called it a game of telephone, and Doug refined the point. It’s not just that the message changes as it passes along, it’s that information gets lost. Details get missed. And the head of sales is left making a judgment call on incomplete data, about deals and about which of their people to trust.
Thang’s favorite bit of irony: he ran forecasts on Excel spreadsheets during his years at Salesforce, the company whose entire product is the system of record. Even there, the number lived in a spreadsheet.
Where it actually breaks down
The people closest to the deal have the best view of it, and the most reason to see it through rose-colored glasses.
The rep is emotionally attached. The first-line manager has strong intel but shares the rep’s upside and downside. So both keep private spreadsheets, take private notes, and don’t push everything transparently up the chain. Why would they? If leadership scrutinizes too hard, the deal gets judged, or worse, the manager’s own gaps get exposed: the POV that started before the economic buyer signed off, the step that got skipped.
Underneath it all is fear. In a lot of orgs, calling a number below quota puts your job on the line instantly. So reps get conservative, hold information tight, and the forecast fills with emotion and politics at every layer. As Thang put it: emotion is something you want in some relationships, not in your forecast.
Culture can change this. At Harness, Doug worked for a CRO who didn’t want optimism or rosy pictures. He wanted the truth, and there was no penalty for calling a number below your quota. That single cultural choice, no penalty for honesty, does more for forecast accuracy than most tools.
But culture alone doesn’t surface the signal. Even with crisp definitions of “commit” versus “best case,” reps put deals in commit that have no business being there. Thang described asking about an $80K “commit” deal: Is the economic buyer engaged? Do we understand the paper process? Do we know the close date? The answer came back: “We’re still working on that, but I feel really good about it.” That’s not a commit. That’s a feeling.
The board makes the stakes non-negotiable
The CRO manages a tight window of accuracy against the board’s number. Land outside it and confidence erodes. And confidence is the whole game. Lose it, and the board starts wondering whether someone else should hold the seat.
Forecast too low and the board questions your belief in the business. Forecast too high and miss, and you’ve triggered a different, worse set of problems. Land more than 10 to 15% off in either direction and it reads as one thing: you don’t understand your own business.
The fix, Doug said, isn’t more feeling. Gut instinct is fine as a starting point, but it has to be backed by what the customer actually did: communication volume, legal red lines, pricing conversations. The trouble is that the only way to gather that today is twelve conversations up and down the chain. The signal exists. It’s just been too cumbersome to collect.
What Airspeed does differently
Airspeed was built AI-native from day one: agents that understand every deal by processing every call, every CRM update, every email, and building context over time. That makes it three things at once: a system of intelligence, a system of conversational capture, and, the part Thang emphasized most, a system of execution. It doesn’t just brief you. It drafts the follow-up, updates the CRM, and coaches the rep on the next best move.
Two analogies landed the point. First, the chief of staff: smart, in every meeting, perfect recall, not emotionally tied to any deal. And it doesn’t just tell you what’s wrong, it fixes it. Second, the casino camera: hundreds of cameras record everything 24/7, but the footage is worthless unless something alerts you that there’s a card counter at table five and a guard five feet away. Most revenue orgs are recording everything and watching almost none of it.
The scale gap is stark. A single average deal might carry 72 touchpoints; across 200 deals that’s over 14,000 data points a quarter. In a typical forecast conversation, a team surfaces maybe 5 to 10% of it.
Take a real-shaped example: ACME Corp, $120K, negotiation stage. The rep says verbal yes, paperwork in motion. A clean commit. Airspeed reads the actual signal: the close date has been pushed three or four times, the economic buyer has been silent for 30 to 40 days, legal never started, and on a recorded call the champion said “I need to check internally on the final steps.” Probability of closing this quarter: 32%. Not a veto. A flag that says look closer, here’s exactly why.
Seeing it live
Alex walked through Airspeed Forecasting doing precisely this. Reps submit forecasts with a close probability attached to each deal and, crucially, the evidence behind the number, pulled from calls, emails, and deal-room messages. Managers get the same introspection across the whole team’s roll-up.
The sharpest moment was “signals your rep has missed.” Agents surface what didn’t make it into the forecast comment. In one case, a “Claude MCP” hard requirement a buyer flagged as critical on a June 29th call that the rep simply never logged. And at the top of the house, a VP-level view that separates the submitted commit from the defensible one: a team rolls up 46, the agent can confidently defend 39, and it shows you the 7K at risk and why.
As Doug summed it up: you’ll never predict with 100% certainty whether a deal lands. But you can give the leader far better data for the judgment call, data that was simply too cumbersome to gather before. That’s the whole point.
Three things to take into your next forecast
- Find a sharper way to interrogate your own number. Stop accepting “I feel good about it” as evidence.
- Know where your reporting chain distorts the number (the game of telephone from rep to manager to CRO to board) and close those gaps.
- Judge deals on customer signals, not internal optimism. What the buyer did beats what the rep hopes.
If your next forecast comes with better questions, less theater, and fewer end-of-quarter surprises, the signal is doing its job.
Watch the full session
The complete conversation and live demo are on demand: Your Sales Forecast Is Lying to You →
Follow Airspeed on LinkedIn to catch the next LinkedIn Live, upcoming webinars, and fresh GTM insights.
More on forecasting from the Airspeed blog:
- Why most sales forecasts are wrong, and how to actually fix them
- Introducing Airspeed Forecasting: signal-powered revenue predictions
- Forecasting you can actually trust
Want to see it on your own pipeline? Book a demo, or switch to Airspeed and get two months free →
