A mash of two SaaS acronyms – CAC being the cost of acquisition and ARPU being average revenue per user; produces a metric which represents the time to payback the initial outlay in acquiring the customer. A simple example, it costs $250 to acquire a SaaS customer via marketing and sales costs (CAC) for a future monthly revenue of $50 (ARPU), in this case it takes 5 months of revenue to repay the outlay (i.e. 5 months of ARPU). The lower the CAC is relative to ARPU, the sooner the customer becomes profitable (assuming cost to serve or CTS is lower than ARPU – if it isn’t perhaps it is time to revisit the business plan).
Within SaaS companies, payback time (i.e. CAC months of ARPU) is a key component of the business’s success. SaaS companies tend to chase growth and customer acquisition in the near term as a pathway to adding long term value. This feature has significant implications on the company’s near term profitability, cash flows and therefore cash requirements – a large CAC months of ARPU results in a large near term expense taking a longer time to repay, therefore requiring significant cash to cover this expense.
So we know that a lower CAC months of ARPU is good, ARPU is what the market is prepared to pay for the product and CAC is marketing and sales costs that go into acquiring the customer. Assuming that ARPU is fixed (the market has spoken, we will pay x and no more), exploring CAC months of ARPU further really comes down to understanding the costs of acquiring a customer at different stages of the company’s growth profile or s-curve – in particular the number of customers being acquired and the costs at the time. Very quickly the s-curve presents a period of low growth as the company starts out, ramping up into exponential growth, followed by a period of reduced month on month growth as it approaches maturity where customer numbers remain relatively constant.
CAC months of ARPU is not static, irrespective of monthly and seasonal variations, it generally expected to trend, what does this trend look like?
In the beginning, only a small number of customers are being added yet there are two key components of CAC in play here. Firstly, awareness of the product is low so there is significant marketing cost that is only bringing in a small number of customers. Secondly, there is often a fixed level of sales costs (not often can you employ part people and associated overheads) – the result is a relatively high near term CAC months of ARPU. Momentum gathers, marketing casts a wider net via feedback from the originally acquired customers – some of this can be attributed to organic growth and marketing lessons are learnt and the money can be funnelled into more valuable avenues coupled with the fact that up to a point the fixed level of sales costs can absorb some of the additional customer acquisitions at no further costs. For example, the fixed salary of one person could as easily cover the acquisition of one customer or (say) 10 customers, CAC per customer is far lower in the second case – both points see CAC months of ARPU reducing from initial levels.
This reducing CAC months of ARPU continues to a point where it starts becoming harder and harder (more costly) to acquire new customers, CAC months of ARPU can begin to climb if the fixed component of the costs cannot be minimised. This often coincides in some relation to the point on the s-curve where month on month growth begins to taper. How quickly this occurs depends directly on how the CAC splits between variable and fixed costs – high levels of fixed costs will see the CAC months of ARPU swing more dramatically than those with higher levels of variable costs.
At the maturing/mature end of the s-curve growth profile is the fixed component of the CAC which cannot be minimised. This is again a period of low customer acquisitions for considerable cost as per the beginning of the growth profile.
Of course this is very stylistic but assuming there is a component that is fixed (i.e. minimum number of sales/market staff) then the CAC months of ARPU over the lifespan of the company’s growth profile follows somewhat of a U-curve – CAC months of ARPU being the highest at the beginning and end of the growth profile, reducing as growth increases and increasing as growth decreases.
Figure 1: S-Curve growth
Figure 2: U-Curve CAC months of ARPU
With all of that laid out, how is CAC months of ARPU used in the financial modelling of SaaS companies?
Given its place of prominence, CAC months of ARPU will most likely be talked about, monitored and reported and therefore is likely to be an output of any modelling. However, it is often also set as an input. Where this is the case, care is required not to simply input reducing CAC months of ARPU, thought is required towards the s and u curves. This approach also requires CAC months of ARPU inputs for each of the periods in the forecast period and estimations of when it will start to cost more per acquisition. This can work as long as there is an understanding of the limitations of the approach.
Clare Capital’s preferred approach is to consider CAC as made up of three components; variable, fixed and semi-variable and report CAC months of ARPU as an output.
The variable component can be considered as the direct marketing spend and sales costs (commission) associated with each customer – this can be a static amount or include a ‘learning’ curve which sees this cost reduce over a period of time to a minimum amount. The fixed component is the minimum costs of market and sales salaries and overheads that are present and will remain irrespective of the number of customer acquisitions. In the simplest example, the initial sales and marketing salaries can absorb far more acquisitions than those that initially occur and are likely to remain towards the end simply to keep stable customer numbers (often includes sales and marketing management). The semi-variable component (or semi-fixed) is reflective that a salaried sales staff member can manage x number of acquisitions per month (per year), any more than x per month and another sales staff members is required, any less and a sales staff numbers can be reduced by one (in reality there is likely to be a delay in this) – resulting in a lumpy semi-variable cost.
The combination of these three components of CAC (variable, fixed and semi-variable) coupled with a correctly employed s-curve results in a CAC months of ARPU that follows the expected profile with a limited number of inputs.
Any questions? Anything we’ve got wrong?
Note: Attached is a copy of the PDF.