automated data affects revenue forecasting process

How Automated Data Processing Directly Affects Revenue Forecasting Processes

It’s Friday afternoon and individual sales forecasts are due before the end of the day. As the sales manager, you know you’ll be spending the weekend massaging the data for Monday’s meeting. You’ve been doing revenue forecasting long enough to know how the numbers are going to fall. You’re hoping it will only take a few hours out of your weekend.

You know that region 1 always inflates its forecast so you’ll cut the projections by 20%. Region 5 seems to be on target most of the time so you’ll leave the numbers alone. With regions 3 and 4, it’s a 50/50 split. Sometimes, they’re over or under, but never accurate. Region 2 is either accurate or under projections.

Monday morning arrives and you feel good about your forecasts. You didn’t touch region 2, but you adjusted region 3 up because you know it’s to close on a big sale in the next 30 days. You lowered region 4 because the majority of its sales were from a sales representative that always overstates his numbers. You aren’t sure the numbers are accurate, but you’re confident that you can explain how you came up with them.

Accurate Revenue Forecasting

According to Gartner, only 45% of companies have confidence in their sales forecasts. Over 50% feel that poor data quality is a primary contributor to inaccurate forecasts. Poor quality data undermine the decision-making process. If the numbers show a result that is contrary to that “gut feel,” people will go with their gut.

sales team doing revenue forecasting
Revenue Forecasting

A recent study found that about half of employees trusted their “gut” over data when making a decision. For C-suite executives and senior management, the gut won 65% of the time despite data-driven insights. It’s no wonder revenue forecasting isn’t accurate if adjustments are made on intuition rather than data.

Start with the Numbers

Sales forecasting is either top-down or bottom-up. Top-down forecasts look at the market size and determine what percentage of the market the company believes it can capture. Bottom-up numbers start with the number of items to be sold, often as a percentage increase over the previous period. In both cases, the number is a target that forms the basis for revenue forecasting. If the numbers appear arbitrary, executives and employees do not value or trust the resulting insights.

Manual Data Processing

How do companies come up with the data to make these targets? They look for external sources that set market size. Then, they try to decide the percentage of market share. With luck, the sales order process or accounts receivable software can indicate the number of items sold and at what price. Getting that information may require an export of data from one system and imported into another. Once the information is available, organizations can “guesstimate” a market share.

The bottom-up approach can use the same sales order or accounts receivable data to determine how many of each product was sold. The company then determines how many more items can be sold in the upcoming month, quarter, or year. If available, organizations can determine the desired percentage of increase over the previous year. Without an established benchmark, the percentage increase lacks the authenticity to be considered credible.

Automated Data Processing

How much of a guess the revenue forecasting is, depends on the starting numbers. If organizations use a manual process, data transfer requires an employee to manually export the data from the accounting package and import it into another system — often a spreadsheet — to manipulate. In some instances, the data may be exported from the spreadsheets into planning software. Each time an employee runs the export/import process, the probability of errors increases.

Automated data processing (ADP) helps maintain data integrity. Information is extracted automatically from the accounting system and fed into the planning software. There’s no need for the interim spreadsheet step and there are no opportunities for human error. Determining a percentage of market share or products sold has more validity when opportunities for error are decreased. Sales staff will be more inclined to provide accurate forecasts if they do not consider the starting numbers arbitrary.

Stop Shadow Forecasts

Shadow forecasts are the spreadsheets or files that salespeople use to track their “real” revenue forecast. These forecasts allow sales staff to run “what if” scenarios and tweak their numbers as the landscape changes. Given the volatile nature of sales, the shadow forecast is probably the most accurate.

Manual Data Processing

Organizations that use CRM or sales planning software often have stages for the sales cycle. These stages are rarely well-defined and are subjective in nature. Sales staff often view updating the software as futile with little thought going into the process.

Exactly, what does “in discussions” mean? Given that many salespeople consider picking a stage the same as throwing darts at a dartboard, it’s no wonder they maintain a shadow forecast. Maintaining a shadow forecast is a manual process that takes time. Defining the sales stages to make them more meaningful may eliminate some of the subjectivity in sales forecasting. Minimizing subjectivity can result in more consistent reporting and more accurate forecasting. However, it’s unlikely to eliminate shadow forecasts if the system lacks what-if capabilities.

Automated Data Processing

What if all data was maintained in a centralized location and access was restricted according to a least-privilege model. With data in a secure location, individuals can access their forecasts anytime. Giving salespeople tools to conduct what-if analysis on their data takes away the need for a shadow forecast. Some may be hesitant to use the system until they realize how much time is saved by not maintaining two forecasts.

Giving sales staff access to data visualization tools makes it easier to see patterns and deviations. Using customizable dashboards sales managers can see changes in a region’s revenue forecasting in real-time. Automated processing of changing forecast data makes it possible for managers and executives to remain informed without increasing the administrative burden on salespeople.

manual processing of data businessman working on a laptop with graphs and charts
Automated data process

Forrester’s 2020 Manager Activity Study found that on average salespeople worked 52.3 hours per week but only 23.3% of their time was spent on direct engagement or selling activities. Automated processing frees sales personnel to spend more time doing what they were hired to do.

Share Business Insights

Without automation, business insights are often based on intuition rather than data. When decisions must be made in a matter of days, waiting a week for the information isn’t viable. The lack of credible data only reinforces the “go-with-your-gut” approach.

Manual Data Processing

Take the sales manager who needed to prepare a revenue forecast for a Monday meeting. Sales staff were given until Friday to provide updated information to their regional managers who probably adjusted the data before sending it to the sales manager.  The sales manager massaged the forecast before presenting it at the meeting. How many people would trust their financial future to manipulated data?

Most organizations do that every time they produce revenue forecasts using manual processes. What happens when a new regional manager steps in and uses different “adjustment” criteria? Or, a manager accidentally reverses two numbers changing the closure percentage from 37 to 73?

Automated Data Processing

Not only can automation improve data accuracy and shorten delivery time, but it can also lead to business insights. For example, a contributing factor to shadow forecasts was the use of “fuzzy” terms to identify stages in the sales cycle. For most salespeople, the stages had little meaning so they picked a stage just to get through the forecasting requirement. Management uses the stages to indicate the probability percentage of closing a sale. The disconnect in providing the data and processing it contributes to the inaccuracy of revenue forecasting.

The sales manager decides to quantify fuzzy terms such as in-discussions. Because the data is stored in a central location, the manager can query the system asking for sales forecast by the opportunity status of in discussions. Those opportunities can be tracked until the sale is made or lost because the data is automatically moved to the centralized data warehouse or lake.

After a few months, the manager starts to see a pattern. An in-discussions status seems to indicate that the salesperson sees a sale as highly likely. After further analysis, it’s decided that the in-discussions status should be weighted to indicate a sale within 30 to 60 days. This decision is based on performance data collected and processed using automated technology.

Chasing a Unicorn

Are accurate revenue forecasts even possible? Or, are they like the unicorn and only a myth?

To deliver accurate forecasts, organizations need credible data that is centralized to create a single source of truth. Automated data processing can help collect, store, and transform data to deliver business insights that are more accurate. Using data that is automatically added to the data store ensures that the most reliable data is being used across the enterprise.

Will revenue forecasts ever be perfect? Probably not. Predicting the future with 100% accuracy has become more challenging with supply chain disruptions, geopolitical realignments, and worldwide health concerns. However, accurate data can help organizations identify where to improve and how to deliver better revenue forecasts.

Delivering revenue forecasts that are 100% accurate is like chasing a unicorn. Delivering reliable revenue forecasts that more accurately reflect performance is not. If your business needs more reliable revenue forecasts based on accurate data, contact us!

Share this on