How to Build a SaaS Cash Flow Forecast

Cash is the lifeblood of a software business. It's the lifeblood of every business. According to CB Insights, it is also one of the top 10 reasons for failure. And while that seems obvious that running out of cash results in failure, what is nuanced is the psychological and decision-impact that running low on cash has on the founder and the organization. "Pennywise and pound-foolish\" is not as antiquated a statement as it may seem. Get low on cash and the exponential forces that were propelling your business forward start conspiring your demise. For founders, pressure builds and the body starts exhibiting Eminem "Lose Yourself" opening line symptoms\...

The wise founder loses themselves in the numbers, not the music. They know the levers of cash generation intimately, and they do not outsource this important function to people that have less to lose and less on the line. Take hold.

The goal: Create a cash prediction model and process that can be a living document and receive updates that make future cash projections more accurate.

How Golden Section can help: The venture partner can help here. Golden Section can help through templates and best practices.

Prerequisites: Note that you do not have to have significant (or any) revenue to build a competent cash flow forecast. But, you will need informed assumptions on the unit economics from your customers.

Steps:

1. Prepare: Get some whiteboard time to map out the unit economics of your business. Usually, this centers around the corporate user in the system, and includes what it costs to acquire them, what and when they pay you, the cost to service them, and how often they churn or upsell. We suggest meeting with a cross-section of your team (get more people in the room than just you) and answer these questions:
   • What does a customer cost? This should include your entire sales team cost, as well as the variable expenses (commissions, travel, 3^rd^ party licenses, etc.). Be careful to fully load these expectations for the 'incremental' expense. Hence, do not bank on getting the next sales employee to come in as low on base as your current team. You need to expect market expenses here.
   • What is the customer payment and terms? This can start with your rack rate. But all good models are driven off of averages. It isn't practical to nail the customer distribution against your service levels in a forecast. Instead, think through averages. For instance, what do you expect the average revenue per account to be and why? Also, what will the average customer do in days payable?
   • What is the cost to service them? This always includes your delivery costs (e.g. hosting etc.) but also includes support personnel, support tools, support leadership, services team, etc. Make sure to be conservative here.
   • Churn or upsell or both? What you are looking for here is the net churn effect. What will your customers (in whole) provide in net churn (upsell revenue from current customers less lost revenue -- usually stated in annual terms). High functioning teams experience negative net churn (i.e. upsell \> churn) but be careful about forecasting that if you aren't already experiencing it. Even if you are experiencing it, you may see it is hard to hold as you scale.
   • What are normal metrics for my industry? You may have historical data to investigate trends. If so, congrats! However, even if you do, we suggest looking at some industry benchmarks to build in an expectation of reversion to the mean in your assumptions. Good sources for B2B SaaS are the KeyBanc SaaS survey as well as the OpenView SaaS Benchmark survey. Of course, Golden Section Ventures has a data set as well.
2. Draft 1: Armed with the data from Step 1, sit down with the template model and build your forecast. This will take some iteration between metric expectations, hiring, cash available, and investment. You should keep metrics in the 'worse than average' category (unless you have a compelling reason). Next, your goal is to moderate growth and hiring to minimize investment needed while still reaching

$5-15M in ARR within 3-4 years.

1. Feedback: Next, send your model out to a few people for feedback. You are looking for diversity of thought in this step. You may want a set of people that accomplish the following characteristics: analytical, creative, process oriented, non-conformists, visionaries, and followers. Ask them these questions:
   • What is good about the model?
   • What is bad about the model?
   • What is missing?
   • What is wrong?
   • Note: Your model should have a link (or note) to evidence for each of the metrics or changes in the model.
2. Final Draft: Incorporate the feedback into the model, but make sure to ensure it is right first. Not all feedback is equal! Once your model is accurate, think through incentives. Who on your team pulls which cost/revenue levers? Make sure they are incentivized to hit the metrics that drive those levers.
3. Iteration Cycle: Designate someone on your team to own the model (but not change it). That person (in the early stage it will be the founder, and later the CFO) should update the model with actuals and report to the relevant team members. On iteration, you are mostly looking at the trajectory of the cost/revenue driving metrics. Are they converging to the average? Are they going the wrong way? These metrics (churn, ARPA, acquisition cost, quota attainment) are key drivers to your future cash.

Troubleshooting

My model shows less than $2M cash needed to achieve $15M in revenue. While this is possible, you are probably proof do you have of top 10% performance? Be careful about that. In fact, go back and do the whole play over.

I don't have sales people and so the quota concept doesn't quite work for me. Understood, that can happen. First, we suggest expecting some sales assistance in the sales process. This may mean a very low quota (or conversely, very high if it is light touch). But if you must, you can remove that part of the model to replace it with a marketing-driven acquisition.