Quantitative forecasting methods
All quantitative forecasting methods follow the same general principle: By looking at a dataset, either historical or current, it is possible to recognize patterns and extrapolate what those patterns will look like if they continue along the same path into the future. Where quantitative forecasting becomes complex is in the finer details:
- What data will be used to build a predictive model?
- What algorithms or formulas will produce that model?
- How near or far into the future will a given forecast look?
Read more below.
It's time for accurate sales forecasting
The most powerful forecasting tool from the world’s largest network of sales professionals.
- Dramatically improve forecast accuracy
- AI-driven forecasting
- Increase accountability across selling teams
- Eliminate inaccurate CRM data
Automated daily sales forecasting
Updated daily based on your data and external network factors that influence your sale.
- Faster predictions
- Reduce sales cycles and analyze scenarios in minutes
- Real-time, buyer-specific selling recommendations
- Increase deal velocity
Eliminate spreadsheets. Automate your CRM.
Eliminate spreadsheets. Eliminate opinions. Get data-driven results with little to no effort.
- Automated data capture
- AI-driven forecasting
- Network enrichment of contacts
- Networked intelligence - unlock external factors that matter
Quantitative forecasting methods
In the simplest terms, sales forecasting has traditionally been a process that relies on historical information to make future predictions. But exactly how sales teams collect and weigh that information has been anything but simple. Relevant information for building a sales forecast can come from a wide variety of sources and methods. For example, experienced, old-school salespeople may predict the number of sales or judge how a deal might move forward based on their gut feeling from initial conversations with a client. Another method that some sellers use is to look at historical sales data to inform the likelihood of certain deals closing depending on what stage they’re currently in.
In reality, most forecasters do not use just one type of sales forecasting. Both opinion-based and data-based sales forecasting methods are utilized by sellers to create sales forecasts. Expert opinion and hard data will always be useful information for any sales organization looking to establish goals and understand what the future might look like. However, though opinion-based methods can give broad insight, data-driven forecasting methods, also known as quantitative forecasting methods, are favored by many for their increasingly complex set of techniques that help forecasters reduce human bias and make decisions based upon data.
With the advent of modern, prescriptive forecasting powered by tools such as Collective[i], sellers don’t have to rely on learning complicated formulas or pouring over detailed spreadsheets. However, having a deeper understanding of the many quantitative forecasting methods that sales teams have historically employed — and their limitations — can help sellers get the most out of their forecasting tools.
In that spirit, let’s dive into the numbers with a closer look at quantitative forecasting methods.
What is quantitative forecasting?
Quantitative forecasting refers to forecasting approaches that seek to make predictions about the future based on data. To understand what quantitative forecasting is, the first step is to understand what it is not. “Quantitative” and “qualitative” are two words that are often seen together and can be easily confused. Qualitative forecasting is the human element of sales predictions; it’s largely driven by expert instinct and opinion. For example, let’s say that the CEO of Accounting Software Company decides to put together a focus group of industry experts to look at its latest accounting software product. The company’s development and marketing teams ask those experts about their opinions on the tool and how well they think it will stand out from competing tools and sell. Based on that feedback, the company can refine the tool and estimate sales — while there may be some valuable insights gleaned, it’s difficult for those opinions to accurately forecast real-world sales.
Quantitative forecasting, on the other hand, focuses on hard data. Instead of using the focus group to predict sales for the accounting tool, the CEO could select one of several quantitative forecasting techniques to generate a report of future sales. When looking at quantitative forecasting methods, it is important to make sure that the method selected is actually based on data, such as historical sales numbers or economic indicators, not human instinct. Quantitative forecasting techniques include using simple math, advanced formulas, and statistical analysis to get an idea of future sales.
What are the different quantitative forecasting techniques?
All quantitative forecasting methods follow the same general principle: By looking at a dataset, either historical or current, it is possible to recognize patterns and extrapolate what those patterns will look like if they continue along the same path into the future. Where quantitative forecasting becomes complex is in the finer details: What data will be used to build a predictive model? What algorithms or formulas will produce that model? How near or far into the future will a given forecast look?
Sales teams can customize their sales forecasts by selecting different combinations of available data, predictive models, and time periods. There are many types of predictive models that analyze data in different ways to create projections. Each sales forecasting tool leans on its own set of algorithms and models to build its forecasts.
Let’s take a look at some quantitative forecasting examples of some of the most common predictive models:
The naive forecasting method is one of the simplest methods out there. The formula is easy: Assume that the next period of sales will be the same as the last period. The period chosen is typically month to month. This method doesn’t require a lot of math, but it’s limited since it doesn’t account for variables. A variation of the naive forecast is the seasonal naive forecast, one of the common quantitative methods of demand forecasting, in which sellers use the same period from the previous year to forecast for the current year — so instead of using December to predict January, the previous January is used to predict the current January.
The run rate method is very similar to naive forecasting with a few key differences. To calculate the run rate forecast, divide the total revenue by the sum of past sales periods. This can be helpful in forecasting sales revenue for specific products. For example, the Accounting Software Company had sold $21,000 worth of one of its products by March one year, and the sales team is attempting to generate a report to show how much revenue is expected by the end of the year. Sellers can see that they made an average of $7,000 revenue per month. There are nine months left in the year, so the run rate model calculates that 9 x $7,000 = $63,000, and sellers can say on the report that for this year, the Accounting Software Company’s projected revenue is $21,000 + $63,000 = $84,000. This method is helpful for setting goals and getting a rough estimate of sales.
Straight-line method or historical growth rate method
Another example of quantitative forecasting is the straight-line approach. This method is fairly simple and commonly used when businesses are in a growth period and anticipating continued increases in sales. The straight-line method is calculated by finding the past revenue growth and applying it to the future. Businesses look at the rate of growth from the start to the present and calculate the average. For example, let’s say that the Accounting Software Company had an average growth rate of 10% in sales from the time of its opening in 2010 until 2020. Using the straight-line method, the Accounting Software Company would apply that growth rate to sales in 2021 and assume that sales would continue to increase. This method does have some limitations since it doesn’t take into account variables such as seasonality or potential short-term economic factors such as a recession.
Trend projection: Graphical method
A variation of the straight-line method, the graphical method of trend projection works by taking the annual sales data and putting it on a graph, then drawing a line through it to visualize the trends. This has the same limitations as the straight-line method since it doesn’t take into account the many variables that affect sales. Also, human bias can easily creep in during the interpretation of the trends on the graph.
Moving average method
The moving average method is similar to the straight-line method, but instead of taking into account an entire set of historical data, it looks at sections (whatever is determined to be most helpful for a particular business). For example, instead of looking at the entire average growth rate for the Accounting Software Company, a moving average would focus on the trends within those years. To find a two-year moving average for this dataset, the model would take the years 2010–11, 2011–12, 2012–13, etc., average them, and compare them. The easiest way to make sense of moving averages is to plot them on a graph to see the trends.
Weighted moving average
The weighted moving average method takes into account that the more recent data is probably more accurate than older data. The Accounting Software Company’s sales from 2010 wouldn’t be too helpful in forecasting sales in 2021 because the market would have shifted, and customers would be making different decisions. To account for that, the formula assigns a numerical weight to the more recent years to increase their impact on the projection, which produces a more accurate forecast that takes the shifting sales patterns into account. This is a technique known as exponential smoothing, and it can also be applied to other methods such as the straight-line method. To calculate the weights, exponential functions are used to assign exponentially decreasing weights over time. For example, if the Accounting Software Company used this method to project sales for a specific accounting tool it’s been selling for 20 years, more weight would be placed on recent years’ sales figures than older figures to account for broad changes in the market over time.
Simple linear regression
This method is more complicated than the previous ones but can result in valuable forecasting data. A model compares one independent variable with one dependent variable to determine the relationship between them. This requires some statistical knowledge, and some programs such as Excel or other forecasting tools can help plot the results of the model. For example, the dependent variable could be the number of accounting products sold, and the independent variable could be the amount spent on marketing the products each month.
Multiple linear regression
Multiple linear regression is a progression of simple linear regression, but instead of comparing just one independent variable, it compares multiple independent variables with one dependent variable. This method is valuable because it accounts for many factors and can provide more accurate forecasts about buyer behavior and future marketing spend.
One of the more complicated methods, the Box-Jenkins analysis uses some elements of the moving average method and is similar to multiple linear regression. It goes through a process of identifying, fitting, checking, and using autoregressive, integrated moving-average (ARIMA) models to analyze and forecast specific time-series datasets. This method works best when there is quite a bit of historical data available to use.
Download this article: Quantitative Forecasting Methods PDF
What is the future of quantitative forecasting?
While these methods and techniques have been used for decades, they all share the same key limitation: The past can never truly predict the future. These complicated predictive models work to iteratively improve the perceived accuracy of sales forecasts, but they still amount to guesswork — very well-informed guesswork, but guesswork all the same. These traditional quantitative forecasting techniques give sales teams goals to shoot for but not much else.
The next step forward for sales forecasting is to leverage both qualitative and quantitative forecasting techniques in the pursuit of prescriptive sales forecasting. Where the goal of all of the above techniques has been predictive in nature — meaning it’s all about trying to project future sales — modern sales teams are turning to tech-enabled sales forecasting tools that don’t just make projections but also provide sellers with actionable insights and help them make decisions about what they do with their time to drive revenue.
The sheer amount of data available within a neural network is greater than any one sales team’s historical sales data. By applying more advanced, machine learning–driven forecasting techniques to a broader set of data pulled from a growing network of companies, prescriptive forecasting tools can get a clearer picture of buyer behavior. Based on those insights, they can also provide real-time forecasts that are not only more accurate but also more practically useful.
Another benefit to this type of modern forecasting is that it reduces the amount of work that individuals have to do since the complex formulas are created and run by the computer, often with minimal human input needed. More sophisticated neural networks reduce the human bias and error rate, leading to better forecasts.
Move past quantitative forecasting methods with Collective[i]
If the above laundry list of statistical models seems overwhelming, that’s because it is. Sellers shouldn’t have to become statisticians to understand the best path forward to meet or even exceed their sales goals. That’s the idea behind Collective[i]’s Intelligent ForecastTM technology, which was developed by leading data scientists and is able to deliver updates, recommendations, risk alerts, and forecasts to help sellers make quick decisions based on good data.
Collective[i]’s artificial intelligence (AI) tech uses machine learning, deep learning, and a neural network to generate these forecasts. Intelligent ForecastTM uses data from a variety of sources, including a vast network of dynamic market data representing buyer behavior and economic factors across entire markets.
Collective[i] also uses historical data unique to your business as well as datasets of online buying behavior to provide sales forecasts that will help you improve your win rates. Our AI provides insights beyond what you can glean from your CRM by combing through diverse, frequently updated, and relevant data.
Instead of getting bogged down with traditional forecasting methods, imagine the possibilities with Collective[i]. Your forecasting should be working for you, not against you, and Collective[i] can ease the burden and take the difficulty out of creating forecasts.
The world changes every day — your forecast should too. Contact us today to get started with Collective[i].