There are many variables that impact a software as a service (SaaS) business. We have all seen monstrous Excel spreadsheets that attempt to forecast SaaS revenue over a three- to five-year period. The open secret, of course, is that Excel models do not work.
Mike Tyson famously said that “Everyone has a plan until they get punched in the mouth.” Similarly, no spreadsheet ever survived contact with reality.
The main problem is that formulas in Excel models are static, while dependencies among parameters governing our business processes are dynamic. They change over time.
To produce a more accurate forecast, instead of a static spreadsheet, one needs a dynamic model of a SaaS sales process that reflects all of its dependencies, feedback loops, and delays. Such a model would act as a “digital twin” of the process and change dynamically in sync with reality.
In addition to simple questions, such as
- How many account executives do we need to hit our revenue targets?
such a model could also provide answers to questions like:
- When can we realistically achieve our revenue objectives?
- When and in what sequence should we be hiring members of the sales team?
- How do conversion ratios and delays in the funnel influence the makeup of the sales team?
- What effect does the product deployment time have over our gross margin?
Time is a common theme in these questions. Humans are very poor at estimating the impact of feedback loops over time. Our estimates tend to be linear, while changes in the loops are compounding exponentially. We are particularly bad at gauging the effect of multi-stage feedback loops that have delays in them, ergo the business cycles and supply chain oscillations. We need dynamic models to help us deal with reality.
System Dynamics is a discipline that can be used to construct such models. It combines the theory, methods, and philosophy needed to analyze the behavior of complex systems over time.
We used System Dynamics to build an interactive model of a SaaS sales process that you can find here:
It is entirely free. No registration is required. Take it for a spin. It may lead you to unexpected discoveries.
It is a common belief, for instance, that the ratio between the number of sales development representatives (SDR) and the number of account executives (AE) must be roughly 3 to 1, that is, there should be three SDRs for each AE. It turns out that there are many valid scenarios where this ratio is reversed. See if you can find them by tweaking parameters in the model. As the saying goes, all models are wrong, but some models are useful.
Here is how a dynamic model can be used to improve the sales process:
- It can be fitted to historical data and used to run what-if analysis, validate business strategy, and produce sales forecasts.
- It can be connected to data in a CRM system and used to generate real-time recommendations, e.g. “it’s time to add a new account executive” or “reduce deployment time to maintain margin.”
Option #1 is descriptive and predictive analytics. Option #2 is real-time prescriptive analytics, the holy grail.