This is the data you should worry about

Bill Kantor
Jun 24, 2021
4 min read

Updated: Feb 11

Big Data Is Overrated—At Least in B2B Sales Forecasting

“Big Data” has become a ubiquitous buzzword. The popular belief is that the more data you have—and the more types you analyze—the better your decision-making. In this view, organizations without massive internal datasets or the budget to buy external data are somehow at a disadvantage.

But when it comes to B2B sales forecasting, that belief is not just exaggerated—it’s often completely wrong.

Let’s explore—and explode—that myth.

TL;DR

Depth beats data quality and breadth in forecasting.
The “basic three” fields are enough for powerful, actionable forecasts.
More data isn’t always better—sometimes it’s just more noise.
Fast time-to-value matters. Don’t let complexity delay impact.
Forecasts should drive action, not just prediction.

The Three Dimensions of Forecasting Data

You can think of the data used for B2B sales forecasting in terms of three dimensions:

Breadth: How many fields or types of data you include.
Quality: The accuracy, completeness, and consistency of the data.
Depth: How much historical data you have available for model training.

Of the three, depth wins. Every time.

Accuracy is nice, but no one—your company included—has perfect data. And breadth? It's a red herring. More fields don't necessarily mean more signal. Often, they just introduce more noise.

Why Depth Matters Most

What truly drives forecast quality is the information content of your data—how much signal it carries about the sales engine’s behavior over time. That signal comes from depth: enough historical examples to learn from, even if the data isn’t pristine. (See our post about that.)

Yes, better quality helps. But if your data is consistently flawed in the same ways, a good model can still learn from it. In fact, large swings in data quality are more damaging than consistent imperfection.

Fixing every data issue is a costly distraction. Depth lets you power through the noise.

The “Basic Three” Fields

For B2B sales forecasting, nearly every CRM system reliably tracks the history of three core fields per opportunity:

Sales stage (or probability / forecast category)
Amount
Anticipated close date

These three fields—though not always perfect—are usually populated and standardized across deals. With a good model, they’re often all you need to generate realistic, valuable forecasts.

Don’t Be Fooled by Extra Data

What about adding activity logs, sentiment analysis, email metadata, call transcripts, or calendar syncs?

In our experience, these fields add more complexity than value. They’re noisy, inconsistent, and often correlate poorly with actual sales outcomes. Complex models built on this type of data rarely outperform simpler models built on the basic three. And when they do outperform, they are fragile. Meaning they worked better last quarter... but not so much now.

Worse, adding these extra fields introduces delays in implementation, extends model training times, and reduces interpretability. The result? Slower time to value.

Occam's Razor wins: the simplest model that explains observed behavior is often the best.

Proof: Forecasting with Just the Basics

Here’s an example of what you can achieve using only the basic three fields.

Below is a 365-day forward sliding-window forecast generated by Funnelcast. Each day over a three-year period, Funnelcast generated a forecast of expected new customer wins for the following 12 months—using only the open funnel at the time.

Blue line: Funnelcast forecast
Black line: What actually happened

Despite being made a year in advance, the forecast performs well. That’s not because of lucky tuning—it’s just good modeling on cleanly defined basics. And it’s not one cherry-picked result. It’s 730 consecutive forecasts, each looking a year ahead. There's nowhere to hide to make the forecasts look better.

Yes, this chart was selected for marketing, but the performance across the full window is consistently strong. That’s the power of a well-built model using focused data.

Why the Basics Work

These fields aren’t magic. They work because they’re consistently tracked, they represent key deal attributes, and they encode meaningful signals about opportunity flow.

All CRM systems capture them. You likely already have them. With Funnelcast, we can connect to Salesforce or HubSpot via API and start generating forecasts in minutes—no custom setup, no costly data wrangling.

Sample reports

See how to sell more.

Try Funnelcast.

Your reports

Forecasts Are a Means, Not the End

Let’s be clear: the goal isn’t just forecasting accuracy. The real value of a forecast is that it informs action—especially action that changes the future.

We help customers not only see where they’re headed, but also focus on the right deals and optimize their go-to-market motion. For example, the gap analysis chart below shows how much of a plan is covered by closed + expected deals, segmented by industry, and what new demand is needed to hit the goal.