Why Your Forecast Feels Off

Has forecasting ever not felt like getting a root canal? If this sounds dramatic, let’s rewind a week and talk about the process of actually getting to the point of calling a number. Some organizations call these “revenue cadences” but those of us who’ve lived it call them a whole lot of wasted time. 

Monday: CRO drops their call and the VP tells the directors exactly how much they’ll need to make up to cover. 

Tuesday: Pipeline reviews, large deal reviews, 1:1s, etc. 

Wednesday: Team forecast call plus more deal status updates.

Thursday: Put your call in your forecast platform. Or your spreadsheet. Or tape it to a pigeon and hope for the best.

Friday: Hound the team to get all the deal notes, stages and forecast categories updated in CRM and update deal amounts.

Saturday and Sunday: “Microadjustments” because your pipeline for the quarter dropped 200k and you have no idea where it came from. 

And so it goes for all 13 weeks of the quarter. 

Why Forecasting Fails (and it’s not your team’s fault)

Leaders hate this process. Reps roll their eyes at the tedium of it. The CRO will end up saying something like “how do we have a $3M gap to forecast and it’s week 11?”

The reason that gap exists isn’t because your team isn’t selling, progressing deals or closing revenue. It’s because the data we’ve all conventionally relied on to build a forecast is subjective, stale and filtered through the lens of people who have a vested interest in optimism.

Can’t blame your reps for this. It’s a culture problem, not necessarily a pipeline problem. And the biggest issue with this problem? It hides in plain sight.

Every rep is working in good faith. Every leader is trying to cut through the noise. The CRO wants to show the board that the number they’re aligned on is one that can be hit. 

But too often, the number can’t be hit. And that knowledge comes too late. 

The Forecasting Question No One Wants to Answer

I’ve asked myself this question a lot over the course of my career: 

“Is it better to aim high and miss badly, or aim low and overperform extensively?”

This is the loaded question RevOps spins around every GTM planning cycle. The one every sales leader has an opinion on. It’s also the one that has no definitive answer. 

And maybe that’s the problem. Forecasting hasn’t evolved because the underlying frameworks are inflexible. The system is too rigid and the visibility is filtered. 

We still treat a forecast ‘call’ like it’s a sacred pronouncement when in reality, a deal will close at or it won’t. 

If it won’t? You need to know that and you need to know why.

The Future of Forecasting is Signal Based

Enter the new world of forecasting. It’s new, it’s radical, and it’s entirely data driven. 

I like to call this new framework “signal based forecasting”. 

Marketers already lean into the signal based approach: web searches, website visits, hiring activity, and intent triggers. Why should forecasting be any different?

Every call, every deal update, every contact added, every amount change. Each one is a signal answering the only question that matters: “Is this deal going to close when and how we say it will?”

The best part of signal based forecasting? There’s no filter. 

Call sentiment, competitor mentions, timeline changes, stakeholder involvement. All of it is right there in the data. By itself, data is just more noise. You need context. You need translation. 

This is where teams are starting to leverage AI. 

The Steps to Unveiling Your Forecast

Step 1: Benchmark against historical trends.

Data driven forecasting starts with the past. Where we are now vs. the same week, same quarter, last year and past years. 

“But wait! Our goal is 25% growth! Shouldn’t we be 25% ahead?”

Yes. But Will you be? 

Unless something has dramatically changed, historical trends tend to hold. Exceptions exist of course (multi-quarter growth surges, new GTM motions paying dividends), but generally, conversion rates are stubbornly consistent year over year. 

If your pipeline doesn’t show growth, don’t fake it in the forecast. Start with last year and build upward, but only if your pipeline proves it. 

Step 2: Validate Deal Health with Signals (Not Hope)

Don’t rely solely on rep confidence, use their confidence as a layer, not the source of truth. 

Ask the hard questions: 

• How has this customer purchased in the past? 
• Do we have an audience with power? 
• Do we have executive alignment?
• Is budget and procurement ready to go? 

Then look at the signals: 

• Call frequency and sentiment
• Email responses
• Stakeholders engaged
• Last meeting date
• Paperwork in hand
• Deal velocity (Multi quarter slips will kill your forecast)

Hedge out risky deals, especially the ones no one wants to admit are slipping. Don’t be a hero.

Step 3: Interrogate Deal Value

Deal value is the most overlooked lever in forecasting. Just like conversion rates, deal size trends stay fairly constant. 

Ask: 

• When was the value set? (If it was by the SDR and it’s closing next week? Trouble.) 
• Has it changed in the last 3 weeks?
• How does it compare to other similar deals historically?
• Are we seeing the “Death by 1000 cuts” as values quietly fall off at the end of the quarter?

Deal size volatility is a forecast killer. Deals bouncing up and down week to week are a warning. Spotting the drops early? That’s the differentiator. 

What Happens When You Get it Right?

Here’s what happens when you move to a signal based forecast: 

• Forecast accuracy within 3% of actual.
• Forecast calls driven by data, not feelings.
• Reps stop playing defense. 
• Managers coach on risk, not drama. 
• No more week 12 surprises. 

The human brain loves a closed feedback loop. Hit that forecast a few quarters in a row? You trust the process, the team buys in, and the board trusts the numbers. 

From Manual Lift to Automated Clarity

You may be here saying “Okay, but that sounds like a lot of work.”

You’re right. I’ve been the RevOps person running these numbers every week. It is a lot of work. But the best thing about the signal based approach? It’s a repeatable methodology that can be enabled. 

• Build dashboards if you must. 
• Build spreadsheets if you have to. 
• Better yet? Let AI do the legwork. 

AI platforms can summarize the signals deal by deal, triangulate the risks and surface the truth before you hit panic mode. Every frontline leader should be running signal based assessments to support their forecast. Every deal, every week. Put that power in your reps hands? You steer the ship on good data, hit your forecast and everyone makes more money. 

Final Word: Ask Your Forecast to Prove Itself

This isn’t about AI hype, it’s about clarity. Your data will never be perfect, but putting two or three humans between data and insights to predict YOUR business will only magnify the imperfections. You can’t afford that uncertainty. Ask your forecast to prove itself. Ask your data to defend its assumptions. The truth is in there. Go find it.

‍