Serverless Computing Provider (FaaS) Financial Model

Description
Serverless computing (or function-as-a-service: FaaS) is essentially the most efficient compute providing service there is. It has been explored as a more mainstream business model after platform-as-a-service and infrastructure-as-a-service. The point is to proivde execution services for functions when called upon and then shut off when not used. It can potentially provide great client savings and help smooth out utilization of the cloud infrastructure and reduce waste.


This was an in-depth study to fully understand the nuance of billing, units, and variable costs. Let's start from the top. I layered in the option for up to six customer types. Each customer type has assumptions for the amount onboarded per month, the monthly retention schedule of each cohort (defined by a percentage remaining after 'n' months of joining) and that is how we know the customer count at the end of each month per customer tier.


I've included best-in-practice modeling techniques to derive the above in the most accurate way possible. Once we know the number of customers that exist at any point in time, we can start driving revenue assumptions. 


There are a few ways to make money and they include (not mutually exclusive):
 - Charging a fixed monthly fee per customer (this would involve having usage limits)
 - Charging a fee per GB/seconds used (you will need to understand how this works to price it correctly)
 - Charging a fixed fee per function executed (usually denominated in cost per 1,000,000 functions executed)
 - Charging per amount of data transmitted per function (if data is being moved out of the servers during execution)


Many of the top service providers in this space will offer an initial base function usage for free and might be something along the lines of the first 1,000,000 requests are free and/or the first 400,000 GB seconds used are free. This is all configurable on a per customer type basis.


I built the calculations in GBs, but you could denominate in any unit size as long as you make sure your pricing matches the sizing accordingly. One of the main uses of this financial model is to understand what the potential return is on infrastructure costs. By adjusting prices, customer growth assumptions, and variable costs, the model will tell you what is feasible.


Some of the general frameworks I've included are capital expenditures for various infrastructure and that will follow a depreciation schedule based on defined useful life. It will also hit the cost of goods sold line item since it is directly related to providing the service. There are also inputs for function usage costs during execution, such as networking and data transmission costs. I've also included a general cost per customer per month input. There may even be the need for customer service scaling and for that the user can define how many customers a given rep can handle. That will drive salary costs for CS reps. This is another item that will hit cost of goods sold.


To account for fixed operating expenses such as general managers, salespeople, legal costs, and any other fixed monthly costs that are not included elsewhere, I have built in a nice OPEX schedule.


For financial analysis, you will find a DCF Analysis, IRR, and option for raising outside funds or a combination of outside equity and inside / operator equity. There are also pro forma financial statements (income statement, balance sheet, and cash flow statement) that automatically update as the inputs are adjusted. The user can choose to include a terminal value as well (the model can be run for up to 72 months, with a dynamic end date). The terminal value is based on entering a trailing 12-month revenue multiple.


Finally, there are key visualizations that make the entire story more digestible and allow for a quicker understanding of the economics as various assumptions are changed.


A similar template that you may be interest in is this data-as-a-service financial model: https://www.eloquens.com/tool/xm9eU21G/jason-varner-smarthelping/data-as-a-service-financial-model-includes-financial-statements


This template is also included in two bundles:
- All Models Bundle: https://www.eloquens.com/tool/P8Y4TX4v/finance/financial-forecasting-models/financial-models-120-useful-and-usable-logic
- SaaS / Subscription Models: https://www.eloquens.com/tool/wmyQI0eE/finance/venture-capital/the-complete-venture-capital-bundle