The accuracy of forecasts affects every aspect of a company, from the most minute details to decisions that may shape the future of the organization itself. When you miss badly, there are consequences. If you’re an enterprise player you could end up like IBM, jettisoning nearly $9 billion of value in a matter of hours after failing to meet earnings expectations. If you’re an organization with unreliable coffers, you may simply cease to exist. But to become a forecasting soothsayer, you have to know when data tells the entire story, and when it needs to be supplemented.

Learn to differentiate between intuition and bias

In business situations, we aren’t always blessed with the gift of definitive evidence, and this is never true more often than in the case of sales forecasting. Forecasting is inherently a mix of art and science, which means it inevitably must incorporate some component of qualitative valuation.

The key takeaway, however, is not all qualitative valuations are created equal and learning to differentiate between them will have important ramifications for the ultimate success of your forecasting. For instance, human intuition is one of our most powerful traits as a species because it allows us to examine patterns and connect the dots in the absence of statistical proof. But, on the flip side of intuition is bias which causes us to favor an established narrative even in the face of contradictory evidence.

Don’t ignore anecdotal accounts…

When it comes to forecasting, regular meetings with your sales team will be among the most important tools for you to rely on. No one knows more in-depth details about the state of their accounts than the reps who work with them, and these insights will prove invaluable in many ways when you are making key forecasting decisions.

An ample portion of these meetings will center on the review of quantitative data, of course, but that isn’t the only reason why they are necessary. Your sales reps will also be able to impart anecdotal evidence that may have an effect on your choices, and it’s important to give them their due. Sometimes, a salesperson might have good reasons to include potential revenue in a forecasting meeting that might not be found yet in the available data.

…but don’t put all of your stock in them, either

That being said, it’s important to remember that including too much anecdotal evidence that doesn’t have quantitative backing can wildly skew your forecasts.

Specific situations can unfold in multiple ways and for myriad reasons, but long-term trends are almost always supported by data. Data itself tells an important story, and frequently ignoring these stories in favor of enthusiastic accounts can lead to trouble.

Also, remember that forecasting is going to be a learning process for anyone, no matter how much experience you have. But instead of using this truth as an excuse to make mistakes, use it as an opportunity to learn from your process and improve your technique. If you incorporate an anecdotal prediction that doesn’t pan out, investigate the causes in detail so you have a baseline of experience the next time a similar issue arises.

Historical performance results are a popular tool for sales forecasters, and for good reason. Data from the past has an important role to play in forecasting, often because established institutions have a tendency to behave predictably. Companies that have existed for a decent amount of time can gather much of their forecasting data from their own historical results, but even startups can find value in examining industry trends and performance patterns of certain markets.

However, markets are constantly evolving which is why it’s unwise to rely solely on historical precedent in your industry. Industries, product markets, and individual companies are subject to the forces of macroeconomics, regulation, and technological upheaval, and therefore these factors must be integrated accordingly.

Understand the forecasting risks specific to your organization

Successful forecasting is all about mitigating risks and knowing when to take them, so it’s pertinent to examine the concept of risk from a more concrete perspective. Every organization is subject to specific endemic risks that can have a massive effect on their ability to forecast with precision. The better you understand them, the quicker you will be able to recognize and account for them.

For instance, you might have an accurate assessment of how many raw prospects you are able to convert in a quarter, yet you might be in danger of consistently overestimating the average deal size due to a specific product quirk. Do your salespeople, like those surveyed across over 500,000 opportunities, have a tendency to underestimate win probability, and time to close? Or, there could be an inherent feature in your sales cycle that occasionally pushes large deals into the next forecasting period. Know your sales process inside and out, and you’ll be able to adjust for factors outside of the normal data stream.