Early Stage SaaS Unit Economics

East Coast investors have a reputation for being sticklers around numbers, and not “buying the dream”, the same way our West Coast brethren tend to. In this piece I’m going to double down on the first part, and continue to strongly deny the second…

One of the most common oversights of a seed stage SaaS company is not thinking about unit economics at the early stages of their business. Far too frequently we see strong founding teams raise a seed and unfortunately fail to close an A because unit economics never came close to working even though they thought they actually were working. These situations are avoidable IMO, and simply requires some conversation around the unit economics framework and theory.

We all know the basic unit economic math. However, it is the theoretical framework that can lead to significant adjustments in the actual numbers, and ultimately a differing POV between investors and entrepreneurs during a fundraise. These adjustments lead to understandably frustrated entrepreneurs who feel like investors are giving them a moving target. On the other hand, investors are discouraged when LTV / CAC goes from the 10x represented in a deck to a 2x after making what are, in their eyes, basic adjustments to the calculations.


There are quite literally hundreds of articles on unit economic math, but for completeness here is a very brief overview.

LTV = [1 Year of Gross Margin $] X [Expected Lifetime of a Customer in Years]
CAC = [Fully Burdened Sales and Marketing Expense in Period X] / [# of New Customers in Period X]

Before jumping into the math and variables above, it’s important to understand the spirit behind the LTV / CAC formula. The formula is a measure of sales efficiency but also (primarily at the early stage) a fantastic tool to understand business model feasibility vis-à-vis potential for future operating leverage. Operating leverage is not talked about enough at the early stage. It simply asks; if I continue to scale my business at the current unit economics, can I have a profitable business at some point in time?

Think of a P&L stripped down to three very basic line items; revenue, variable expenses (the expenses captured in an LTV / CAC formula; gross margin in LTV, and S&M in CAC), and non-variable or fixed expenses (the expenses NOT captured in LTV / CAC e.g. rent, overhead, back office salaries, etc). If your LTV / CAC equals 1x, your business will never enjoy operating leverage (make money) because the gross margin is only enough to pay for the variable expenses associated with obtaining that revenue, and not the fixed expenses of the business. In other words, the gross margin only covers the S&M expenses associated with obtaining that margin (CAC), and not the rent, overhead or back office salaries.

Generally speaking, investors look for an LTV / CAC ratio of greater than 3x. At 3x, you’re making some money on those customers that you acquired and more importantly you’ve created a product and paired it with an acquisition strategy that fits and will hopefully scale. Your gross margin pays for its CAC and contributes further to the bottom line (net income) of the business.


Now, let’s dissect the variables of the formulas because this is where the miscommunication often occurs between investors and entrepreneurs.

LTV = [1 Year of Gross Margin $] X [Expected Lifetime of a Customer in Years]

1 year of gross margin or 1 year of revenue minus COGS

There are two types of revenue; recurring and non-recurring. For your LTV analysis, you’ll want to only use the recurring piece of the contract as captured in your ACV (annual contract value). If your contracts are annual recurring for $1M with a $100k up-front onboarding / professional services, you’ll want to use your $1M, and dis-include the $100k (note that some businesses have very high margin on their professional services, in which case you can make the argument for adding that one time gross margin to your LTV if that’s the case for every customer). The ACV that should be used is the median number across all current contracts. If there has been a consistent rise in ACVs over the last few quarters, there is a case to be made for skewing larger as opposed to a pure median. However, the average number is almost always skewed by a few large deals (particularly at the early stage) that are not representative of the typical deal you’re signing.

COGS is where a lot of mistakes can be made. For a SaaS company, COGS should include any expenses that are variable and go into the creation, onboarding, maintenance and support of the software. This includes hosting, 3rd party licenses, data agreements, and headcount around deployment, maintenance and customer support. The mistakes made here are generally around the headcount.

LTV = [1 Year of Gross Margin $] X [Expected Lifetime of a Customer in Years]

The expected lifetime of a customer at the early stage cannot be calculated with just math because your total operating history is hopefully shorter than your expected customer lifetime value. That said, the basic math formula (the non-basic and right way to do this is by looking at monthly customer cohorts + churn) is; 1 / annual churn % = lifetime in years. If you’re churning 20% of your customer base per year, that means it’ll take 5 years to go to zero, or a 5 year customer lifetime. What this does not mean is that if you’re a 2 year old SaaS business signing annual contracts with 5% churn, you have a customer lifetime of 20 years. In my own analysis I use 4–5 year lifetimes as a proxy for a healthy early stage company with annual contracts. If a company is purely SMB, or routinely signs longer 2–3 year contracts, the lifetime will flex very slightly down and up from there respectively.

The logo churn rate is the best indicator for an investor at the early stage to estimate customer lifetime. The formula here is; # of lost customers in period / # of customer at beginning of period. All things being equal, I have to assume that going forward all of your customers (across verticals and sizes) are going to churn at the same rate. Said differently, I cannot throw out a 15% logo churn period from two months ago simply because they were SMB or from XYZ vertical which you’ve recently learned is no longer relevant. A caveat here is that if there is a substantial amount of historical data (at least 12 months) that can prove churn rates really do vary across groups, adjustments can and should be made to the LTV model.

The other common mistake we see made in this analysis is using negative revenue churn to justify a very long customer lifetime. This is apples to oranges. The argument is that if upsell is high enough in a cohort to have the offsetting effect of zero churn in a period, customer lifetime should not be negatively dinged. The way we’d run the analysis is a shorter lifetime in years, with a higher ACV driving the numerator in the LTV formula.

Logo churn is far more telling than revenue churn at the very early stages. Logo churn means a customer didn’t work out and proactively decided to stop using the software. As an investor, it leads me to question whether or not the software works for that specific vertical or customer size. Is TAM inherently limited or was this just a bad individual customer experience? Oftentimes it’s the latter and the customer experienced early product bugs or the pitfalls of a constrained support org — things that are serious, but fixable. In any event, it’s a good idea to proactively address these potential concerns, and ideally with referenceable current and churned customers.


CAC = [Fully Burdened Sales and Marketing Expense in Period X] / [# of New Customers in Period X]

The cost to acquire a customer (CAC) is straightforward. CAC equals all of your sales and marketing expenses in a period divided by the number of new customers that those costs helped to sign up. The expenses include marketing, advertising, events, content, and fully loaded headcount for S&M (benefits, commissions, etc). Every organization is structured differently, but if the individual has a hand in anything related to the funnel, they should be included. Marketers, content creators, SDRs, sales engineers, sales reps, etc.

One thing that is an almost certainty at the early stage is that your CAC will increase over the short term. This makes sense because you don’t know who your successful customers will be, your keywords are not optimized and your sales organization lacks late stage sophistication and efficiency. Below is my attempt at illustrating the CAC trend of an average company throughout its full lifecycle. [Disclaimer 1: Trend lines are dramatic to illustrate the path. Disclaimer 2: The slopes and relation of the peaks and valleys vary across companies.] One of the jobs of an early stage investor is to try and estimate how high that first peak will go. It is at that point of repeatability in the sales process that a company starts to see a decline in CAC, and usually a very solid growth equity investor comes in to pour gas on the fire.

Given this trend and expectation on behalf of investors, it’s worth visiting the LTV / CAC target of 3x again. All things being equal, if CAC was the only variable that we expected to increase over time, we probably don’t write the investment check. A great SaaS company will also be able to tell the story of rising ACVs and GMs very clearly, and ideally with few data points. This can come by way of new features, selling upstream or simply increasing prices.


The reality is that no one knows what the unit economics will be until there is a product out in the wild that is being sold. Bill Gurley said it well in this post on unit economics from 2012. “It’s at best a “good guess” about how the future will unfold. Businesses are complex adaptive systems that cannot be modeled with certainty. The future LTV results are simply predictions based on many assumptions that may or may not hold.”

With that said, we’re at a point in time where there are enough comps in the market for new founders to build unit economic models with their best guess of forecasts. Are you selling to SMBs? You’re probably looking at smaller ACVs with higher churn. Are you building a product to dethrone a current incumbent? You’ll likely get similar ACVs with higher CAC initially as you attempt to displace their solution.

As an East Coast investor / stickler for unit economics, I find founders show best when they enter early stage conversations armed with this type of market data. While it’s absolutely not the most important topic during a fundraise, knowing how and why the formulas exist lend to a productive conversation between founders and investors.