Determining LTV:CAC in the age of land & expand
B2B SaaS companies are increasingly deploying land & expand distribution strategies. While land & expand works well for customer acquisition, it can create headache for one of SaaS’s favorite metrics.
The original LTV:CAC formula doesn’t hold up for companies with land & expand motions. And the alternative method suggested of using a discounted cash flow model with assumptions for growth rates, discount rates, and adjusted margins for account management costs isn’t operationally feasible. So rather than calculating LTV:CAC at the customer unit economic level, calculate it at the $1 of ARR unit economic level. And tack on one-time gross profit / loss from professional services.
LTV:CAC is one of the most powerful and most important metrics for SaaS companies. It encompasses all aspects of your P&L into a single metric, and it is the only single metric I know of that tells you whether your entire SaaS business is designed in a way that can turn a profit. The first I read about it (and a lot of other SaaS economic concepts for that matter) was at David Skok’s blog For Entrepreneurs. I believe he may have originally devised the concept - impressive.
From an economic standpoint, a SaaS customer is a series of recurring payments over a period of time until the customer churns. In theory, a customer could never churn, so you would end up with payments to infinity. There’s a term for this in finance, it’s called a perpetuity, and has a very simple formula to determine its present value (“PV”):
P = value of payment
r = discount rate
We can use this exact same formula to determine the value of a SaaS customer, which we call the lifetime value (“LTV”):
The next step is to compare the LTV of a customer to the cost you incurred to acquire that customer (customer acquisition cost or “CAC”). A result of 3x or more typically indicates that your go-to-market model will eventually lead to profitability and may be ready to scale (you should also consider payback periods - a post for another time). Below 3x indicates that you're paying too much to acquire a customer relative to the gross profit dollars that customer will provide in return, and your company's go-to-market model (customer acquisition, margins, churn, and/or pricing, etc.) needs to be refined. More on this later.
This all makes sense. So what's the problem?
This formula holds up well for companies whose customers buy once at a set price (i.e. the arrangement with your customer will remain the same from signing to churn). The real world, however, is more complicated and a lot of value is being generated by SaaS companies through expanding revenue streams from an existing customer (more products, more seats, more usage, etc.).
Consider the following example: Customer A is a simple buy once SaaS customer. You pay $28,000 to acquire Customer A who pays you $20,000 per year. Customer B is a land & expand customer who signs up initially on the same terms and cost as Customer A, but then expands revenue 20% per annum. Your company's customer churn rate is 20%, so both customers have a 5 year expected life. And your subscription gross margin is 75%.
Here I calculate the LTV:CAC using the perpetuity formula in column I, and again by adding up the sum of years 1-5 in column J. Both methods yield the same result for Customer A, but yield different results for Customer B. You'll notice that the perpetuity formula's result is actually incorrect for Customer B. This is because the LTV and CAC for Customer A are fixed at signing, while the LTV and CAC for Customer B evolve throughout the customer's life.
This is a known issue with the original perpetuity methodology, and in a land & expand scenario conventional wisdom says to calculate LTV for Customer B using a discounted cash flow ("DCF") model, burdening gross margins with account management and expansion sales rep costs (to cover the additional CAC), and incorporating a revenue growth rate and cost of capital discount rate. From an academic perspective, provided you can make some relatively accurate input assumptions, this approach is sound. But as an operator with 100 other things going on, it’s simply not feasible to spend hours refining inputs, running, and eventually re-running DCFs for your customers. There isn’t enough time in the day.
An alternative calculation methodology
How else then could we determine the LTV:CAC in a land & expand scenario? Getting back to basics, the whole idea behind LTV:CAC is unit economics. Did your company pay a reasonable amount of money to acquire your customer relative to the cash flow that customer will provide before they churn. And as we defined in our LTV and CAC formulas above, the “unit” in unit economics is the customer. But when the journey of each customer can be unique, we either need a unique model for each customer, or we need to look for an alternative “unit” for the unit economics analysis.
Rather than defining the unit economics at the customer level, what if we determined the unit economics for $1 of ARR? Looking at the unit economics of $1 of ARR will still inform on the viability of our go-to-market model (i.e. did I spend a reasonable amount of money to acquire $1 of ARR relative to the cash flow that $1 will provide before it churns). And because $1 of ARR is the same (on the revenue side) regardless of whether it comes from a new logo, expansion, upsell, or whatever else, we can now accurately and easily determine LTV in land & expand environments, as follows:
What about CAC? Now that we’re defining the LTV on a unit economic basis of $1 of ARR, we also need to define the CAC in terms of $1 of ARR. So our CAC becomes the CAC / ARR ratio. The CAC / ARR ratio is a useful metric for analyzing Sales & Marketing efficiency, and is calculated as follows:
S&M expense in this formula should only be related to acquiring new/expansion/upsell ARR, not renewing customers. You also need to determine the period for which you’ll compare S&M expense and ARR acquired. Some companies like to compare prior quarter S&M expense to current quarter acquired ARR. Depending on your sales motion/cycles, it may or may not make sense to offset. You should make sure, however, that the inputs used in the LTV and CAC components of the formulas are consistent durations, e.g. one year, one quarter, one month, etc.
This method may not be as academically sound as using a DCF, but my guess is your results will be more accurate as you’re not required to make any estimates. And the calculation is very easy to maintain. You can either apply this to your whole business for a blended rate, or as granularly as you want provided you track dollar gross churn, CAC / ARR ratio, and potentially subscription gross margin at the granular level. For companies who operate in various business segments (e.g. mid-market and enterprise), I’d recommend tracking LTV:CAC separately for each. See the below example for the formula in action:
What about professional services (or non-subscription) revenue?
It is not uncommon for companies to sell professional services ("PS") for implementation or training along with a SaaS subscription. How should PS impact our LTV:CAC formula? First, let's look at the gross profit or loss (not just revenue) generated on PS. We then have two options:
- include PS profit/loss as a component of LTV; or
- include as a component of CAC
I prefer to include PS gross profit/loss a component of LTV because the principle of LTV:CAC is to compare the top half of your P&L (through gross profit) to the bottom half (operating expense). So your LTV formula becomes:
So you have your LTV:CAC result, now what? The next step is determining the right LTV:CAC ratio for your company. 3x may not be enough, or it may be plenty. I’ll write another post on this in detail, but the idea is that the LTV of your customers should cover your entire operating expense base. CAC represents the S&M portion of your OpEx base, so if S&M is 33.3% of your total operating expense (i.e. S&M, R&D, and G&A), then 3x is the right number for you to break even. So you’ll want to be above that.
You can also check out how your company is doing relative to other B2B SaaS companies at benchmarkhq.io, and can slice/dice the benchmarks however you want (ARR size, market segment, distribution method, profitability, industry, etc.).