Table of Contents
In a previous post, we examined different types of financial models for SaaS Companies and the three main model types:
- Operating models
- Forecasting models
- Reporting models
Of these model types, companies most often struggle with finding the best way to forecast SaaS revenue. Financial and workforce processes are often siloed and each department works offline on its own plans. As a result the entire planning cycle of the organization is prolonged, and the enterprise cannot get a true sense of the current state or the future state of the business.
Oftentimes, a single modeling approach may not be the best way to predict SaaS revenues, Below we break down 4 common methods that can be used to help forecast revenues for your SaaS company.
Quick Refresher: What is a revenue forecasting model?
Typically, a revenue forecasting model takes a snapshot of a companies current data in order to create a model that predicts revenue based on current trends. These models can consist of many different formulas stemming from simple linear models with singular predictors to multiple regression models with a large number of predicting variables. These formulas all rely on an accurate source of data in order to make a prediction that is up to date with company actions.
How do you forecast SaaS revenue?
Although each of these models may be appropriate some of the time, using a combination of these models can yield a more complete forecast of your revenue. Using forecasting techniques such as sales pipeline, price-volume assumptions, as well as sales capacity, will provide your model with up to date information that will improve decision making.
With these four techniques your models will provide your sales teams with clear forecasts of future revenue that uses all available information in order to make informed decisions.
1. Forecasting Revenue Using Sales Pipeline
Leveraging pipeline information is a key component for building a near-term forecast model. For companies that use a customer resource management (CRM) application, such Salesforce, this can be relatively straightforward to maintain. Middle or late stage opportunities are the best indicator of new deals and the associated revenue streams and future revenues.
Although this data can be a valuable indicator for the next few months, it’s probably not useful for long-term projections. Many SaaS companies use the sales pipeline to forecast the next quarter, but not much beyond that. Hence the need for supplemental SaaS Revenue models. Source
2. Forecasting Revenue Using Price-Volume Assumptions
SaaS revenue forecasting is frequently modeled based on selling price and deal volume. The selling price, or average contract value (ACV), may be an assumption or based on prior sales.
These models typically allow users to enter a number of expected deals in a given period. Entering the number of deals is typically at the monthly, quarterly or even yearly level. Entered as annual numbers along with additional assumptions (such as seasonality and linearity) can spread across the months. They may also be broken down by product, sales region, country, etc.
This model is useful for forecasting many quarters or even years into the future. As such, it can model revenues well beyond where the pipeline data is typically insightful. Because of this, many companies refer to this as their long-term forecast.
This type of model is easily updated or modified, depending upon data input granularity. The inputs for number of deals and ACV are entered in a simple spreadsheet-like data entry form.
A variation on this technique is to model the change in price and/or volume as a percentage of previous entries. For instance, a user may enter an assumed number of deals and ACV for the remainder of the current year. Assumptions (such as a 20% growth in number of deals and a 5% growth in ACV) can calculate the price-volume for the subsequent years.
3. Applying Seasonality and Linearity Assumptions
As mentioned, models often factor in seasonality and/or linearity assumptions. In these cases, a user may enter the expected number of deals for a given year. With that, the deals by month are spread by applying the seasonality and linearity.
Seasonality typically represents the distribution of deals across the four quarters of a company’s fiscal year. Almost every business has some degree of seasonality effects. There might be more sales near the end of the year or a drop in business in the summer months. Likewise, there could be other similar variations from quarter to quarter.
Linearity can represent the reality of a big push to close deals at the end of each fiscal quarter. A distribution of deals closed within a quarter might look like this:
- First Month : 20%
- Second Month 2 : 20%
- Third Month 3 : 60%
4. Forecasting Revenue Using Sales Capacity
The third common method for forecasting SaaS revenues is based on the numbers of salespeople and their expected revenue. As with the price-volume model, this can leverage past data to make educated assumptions about expected revenues generated per sales rep.
In the same way the price-volume model makes assumptions about the ACV and the numbers of deals, this model makes assumptions regarding revenue per rep and the numbers of reps planned.
This model can also incorporate assumptions around percentage changes in future years. It can also apply seasonality and linearity assumptions similarly to the price-volume model.
We have described this model as another method for forecasting revenues. In this case the number of reps and revenue per rep are inputs to the model. However, the inputs and outputs can be reversed. For example, in a price-volume approach, the output could be required numbers of salespeople to achieve those revenue numbers.
Hybrid Forecasting Model
As previously mentioned, a pipeline-based projection of expected revenues might be best for the next few months, but its efficacy drops off rapidly in the long-term. Subsequently, a hybrid model that leverages the pipeline and then transitions to a price-volume and/or sales capacity model might be the best approach.
We have seen customers apply a simple approach such as “the next three months are always using the pipeline, then we switch to only using price-volume modeling thereafter.” Another variation is to use a blending approach where for every given month, the customer can define which combination of models to use with a weighting applied to each model. This can provide the user with visibility into the projected revenues for each model simultaneously, and allow them to modify the relative weightings based on these projections.
Want to Learn More About Revenue Modeling for Your SaaS Product?
In order to be successful, businesses must have a clear set of goals shared by the entire business that are both challenging and obtainable. When revenue forecast modeling is not done on a shared and collaborative platform, these goals can be lost, changed or simply incorrect. This can lead to complacency or panic depending on the shortcomings of siloed revenue forecast models. Often as sales teams and businesses grow, these problems are only compounded.
In order to ensure continued transparency while promoting revenue growth and see what QBIX and Workday Adaptive Planning can do for you and your business.