# SaaS Metrics: Rethinking Customer Churn Rate & LTV/CAC

In 2018, Loomly experienced a 600%+ year-over-year revenue growth rate and passed the $1M ARR mark. This scaling process has provided our team with a nice opportunity to work on real-world data points, truly internalize what each SaaS metric means and figure out how to leverage it to continuously improve our growth engine.

Thanks to the exhaustive literature available online (from The Angel VC, Clemnt, Saastr, For Entrepreneurs, Tom Tunguz, OpenView and ChartMogul to name a few), wrapping our heads around most Key Performance Indicators has been a pretty smooth process. Except… for two metrics that took us down the rabbit hole: **CCR** (Customer Churn Rate) and **LTV/CAC**.

The more articles, studies and benchmarks we were finding, the more obvious it was appearing that we were not seeing the whole picture. What started as a simple “how are we doing in terms of X” management question, opened the door to a deeper intellectual quest to understand the very golden ratios against which we are all measuring ourselves.

Thanks to the invaluable guidance and hands-on contribution of Luc-emmanuel Barreau, Partner at Red River West, we have been able to formulate hypotheses, test them, and draw practical conclusions about these two SaaS KPIs. Here are our counter-intuitive findings.

**Acknowledgement**: as a sanity check, we submitted this article to subject matter experts in order to challenge — and consolidate—our reasonings and findings. We would like to address a warm thank you to Tom Tunguz, Christoph Janz, Clement Vouillon, David Skok, Kyle Poyar, Sara Archer & Ilia Markov for their incredible feedback in this process.

*TL;DR **our main takeaways are:*

*Under certain circumstances, a higher growth rate will result in an apparently higher customer churn rate.**A SaaS business may be perfectly viable with a LTV/CAC ratio lower than 3 depending on your cost structure (and CAC Payback Period).*

# Customer Churn Rate: The Underrated Impact Of Exponential Growth

Stating the obvious: **churn influences growth**.

However we look at it (Customer Churn Rate or Revenue Churn Rate), the faster we lose business, the slower we can grow. And vice versa. These great resources (and many more) cover that topic extensively.

We can also all agree that some **structural factors influence churn**, such as:

*Stage*: younger companies tend to operate with higher churn rates than more mature businesses. According to the folks at Baremetrics (and their Open Benchmarks), this is related to the process of reaching product/market fit:

Most early-stage SaaS companies I’ve observed typically have churn around 10–15% for the first year as they work out exactly what their product needs to do, then they’re able to reduce it pretty quickly.

*Target Market Segment*: smaller customers (i.e. accounts with lower ACVs) tend to churn at higher rates than larger ones. The following table by Tom Tunguz provides a solid benchmark of churn rates by segment:

All of this makes perfect sense and is in line with what we have observed at Loomly and speaking with other SaaS startups.

However, one thing was not feeling right: while we were over-indexing in terms of customer satisfaction (based on direct feedback from users, a 70+ NPS and really nice online reviews), we were not “through the roof” in terms of Customer Churn Rate (i.e. we were “only” in the middle range of the above benchmarks, not killing it).

The “aha!” moment came when we stumbled on this brilliant piece by CatchJS’s Founder Lars Hiller Eidnes:

It’s counterintuitive, but it’s a statistical fact: [monthly CCR]* actually

tells you nothing useful about churn, but really relates to the age of the subscriptions you have. It will in most cases go down on its own, and, absurdly, the only way to keep it from going down is to have very high growth. So the number will literallyonly look, and optimizing for it will be directly counter-productive. The error here is a simple statistical mistake that is easy to make, and luckily also easy to understand and avoid.badif your business is doing extremelywell

**calculated as “Customer Lost During The Month / Customers At The Beginning Of The Month”*

In other words: not only does churn have an impact on growth, but growth can also have an impact on churn! Or, to be more specific, acceleration (i.e. the growth of growth) can have an impact on churn. Needless to say that with a 7x year-over-year revenue growth rate (and a commensurate scale of our acquisition efforts), this realization piqued our curiosity.

Here is the kicker:

The problem is, your customer is not equally likely to cancel their subscription at any time. Most likely, you have a situation where the drop-off in customers is higher in the first few days than it is later.

Eureka: armed with our ChartMogul data, we performed a cohort analysis of the Loomly customer base, and indeed found out that our monthly CCR profile was ** not** linear over time. Instead, it did turn out to be steeper in the first few months, then slowing down (with a punctual peak at 12 months due to annual contract anniversaries) before bouncing back a bit towards the end when the number of subscribers in the cohorts was getting smaller.

The following graph is a simplified way to represent our findings (built from made-up data, not actual Loomly data, for simplicity and clarity purposes):

From the chart above, it is pretty obvious that, **for a given cohort**, CCR actually decreases over time: it is going down (until *Month 9) *and then plateaus (until *Month 36*). In other words, even without any Customer Success effort to prevent churn, churn seemingly “improves” over time.

This is why the million-dollar question actually is: what happens when we start piling up cohorts on top of each others? More specifically: what happens when we start piling up bigger cohorts on top of each others?

Well, to figure out the answer, we did exactly that (we piled up cohorts on top of each others):

- First, we assumed the following monthly CCR profile for a standard cohort:

2. Then, we built a table allowing us to see how the number of customers would evolve if we piled up the exact same cohort every month on top of all previous ones, for a year:

3. This allowed us to calculate an aggregate CCR for each month:

4. Then, we injected the notion of **exponential** growth into the model.

Here, we need to take a moment to make sure we are all on the same page, as this is is a very important nuance, crucial to understand what follows:

- The month-to-month growth rate we are using is
**the growth rate of the number of acquired customers**. - In other words, a 0% growth means that we acquire the exact same number of customers every month (and
that we acquire 0 customers every month); similarly, a 10% growth rate means that we acquire 10% more customers every month than the previous month.*not* - In other words, we are talking here about
**acceleration**(increase in speed of growth) rather than growth itself.

Here is what the table from *Step 2* above now looks like with a 10% monthly growth acceleration rate:

5. All of the above leads us to observe how the monthly CCR evolves when the month-to-month acceleration rate varies, through the following chart:

This chart teaches us that:

- Even with a 0% acceleration rate, the aggregate monthly CCR plateaus at 6%, instead of 3% at the cohort level, i.e. 2x as much.
- A 20% acceleration rate results in an aggregate monthly CCR plateauing above 8%, i.e. 5 percentage points above the cohort benchmark.
- A 30% acceleration rate drives the aggregate monthly CCR to plateau at 9%, i.e. 3x the cohort value.

This is happening because, with that specific churn profile, acceleration entails an ever-growing proportion of younger cohorts, which individually yield a higher CCR, and therefore inflate the aggregate CCR.

The takeaway for SaaS startup founders and investors is very crisp: when churn is not linear over time and looks like the above profile, **acceleration rate should be taken into account when interpreting aggregate CCR**. Said otherwise: all else being equal, with this type of churn profile, when you start to grow faster, expect your CCR to go up.

From a practical standpoint, this has two main implications:

- On the short term, for management purposes:

- In order to understand and improve churn, analysis is more effective when performed on a cohort basis than on an aggregate basis. If anything, a cohort analysis will reveal exactly
*where*(or, should we say,*when*) the leak in the bucket is, rather than simply measuring the hole. From there, you can dive deeper and investigate*why*customers are churning: for instance, depending on whether churn can be correlated to customer specifics (segment, use case, job title, etc.), to a poor onboarding experience, or to a foreign factor (client-side turnover), then the relevant actions will differ. - When benchmarking aggregate monthly CCR— be it against competitors or industry standards — churn profile and acceleration must be taken into account. After all, if an early-stage company scales 7x (with a steep acceleration) over a specific time range, while another more mature business grows less than 20% (at a steady rate) over that same period, but both have the same CCR, they should probably not be considered to have been performing equally in terms of churn (to be clear: the former clearly outperforms the latter).

2. On the long term, for forecasting purposes:

- If exponential growth is part of the plan, then it may be wise to bet on conservative churn rates.
- Conversely, as a business reaches a certain scale and acceleration slows down, CCR should be looked at more optimistically than while in hypergrowth mode.

If you enjoyed this first section, then definitely keep on reading. It gets even better as we share our findings about one of the most popular SaaS metrics: the LTV/CAC ratio 👇

# LTV/CAC Ratio: The Overlooked Role Of Cost Structure

We are taught to believe that the LTV:CAC ratio is magical and helpful. It can also be confusing as shit, and a lot of smart people have written about its many complexities and nuances. Adding to the confusion, there are unhelpful benchmarks that everyone cites. Why is a 3x LTV:CAC ratio the appropriate benchmark? No one knows. It just is.

Unbelievably, this is true: while you can find countless resources stating that a SaaS business is not viable unless it operates with a LTV/CAC ratio greater than or equal to 3, not a single one explains why.

So, why don’t we start with what we know and work from there?

According to David Skok:

In the first version of this article, I introduced two guidelines that could be used to judge quickly whether your SaaS business is viable.

The first is a good way to figure out if you will be profitable in the long run(…)

Over the last two years, I have had the chance to validate these guidelines with many SaaS businesses, and it turns out that these early guesses have held up well. The best SaaS businesses have a LTV to CAC ratio that is higher than 3, sometimes as high as 7 or 8. (…) However many healthy SaaS businesses don’t meet the guidelines in the early days, but can see how they can improve the business over time to get there.

(…)

I should stress that these are only guidelines, there are always situations where it makes sense to break them.

If LTV/CAC is a predictor of the viability of a SaaS business in the long run, it means that it has to do with how a company operates and, ultimately, generates profits.

Back to the drawing board (in our case: a spreadsheet), we created a simplified, fictive company, which:

- Operates with a standard 80% Gross Profit Margin Rate.
- Runs on a cost structure on par with industry benchmarks i.e. a relatively even breakdown of expenses between S&M, R&D and G&A (33% each).
- Serves exactly one client over the course of three years, which yields a revenue of $1,000 per year (with no additional costs associated with the management of that client beyond the first year of service, i.e. no new S&M, R&D or G&A expenses in Year 2 and Year 3 in our model below).

Guess what! In that case, it is correct that a LTV/CAC ratio equal to 3 is the tipping point of profitability:

This makes sense:

- During Year 1: you repay your S&M expenses (CAC).
- During Year 2: you repay your R&D expenses.
- During Year 3: you repay your G&Q expenses.
- During Year 4: you start turning a profit.

That is the reason why LTV/CAC has to be higher than 3 for a SaaS business to be viable on the long run.

You can probably imagine our reaction when we received the following email from David Skok himself:

I should own up and tell you that the reason people believe the number should be greater than 3 is because of me. I was the very first person to start writing about LTV and CAC in the early days of SaaS (2008, in this post), and I put out that guideline, and it was adopted as a universal norm. I think the reason it got so widely adopted was because it actually made good sense. And you are correct to have done the breakdown in the cost structure of a SaaS business to validate the 3x ratio.

But what happens if you serve that exact same single customer, with a slightly different cost structure?

For instance, at Loomly, we:

- Operate with a higher-than-average Gross Profit Margin, close to 90%.
- Run on a very low cost structure (as a distributed team, we have extremely low overhead costs), with S&M expenses about equal to all R&D and G&A expenses together.

This changes the game, as it brings down the tipping point of profitability to a LTV/CAC of 2+ rather than 3+:

In other words, a SaaS business with a lighter-than-average cost structure that has a LTV/CAC ratio of 2, can be as profitable a SaaS business with a standard cost-structure that has a LTV/CAC ratio of 3.

Here again, the takeaway for SaaS startup founders and investors is very crisp: **cost structure should be taken into account when interpreting the LTV/CAC ratio. **Simply put, the lighter the cost structure, the lower the LTV/CAC ratio has to be for the business to be viable.

On the playing field, this has two practical ramifications:

- For entrepreneurs: a LTV/CAC ratio lower than 3 does not necessarily mean that your business is doomed (particularly if your CAC Payback Period is lower than 12 months). Yes, there may be room for improvement in terms of ARPA, CCR and CAC. However, depending on the typology of software & services your company delivers, and the market segment to which you deliver them, optimizing cost structure may present a complementary avenue to reach profitability. Of course, conversely, even with a LTV/CAC ratio higher than 3, your business could lose money if you operate with contained S&M costs (i.e. a small CAC) but high R&D and/or G&A expenses.
- For investors: whether you firmly believe that a business is only viable with a LTV/CAC ratio greater than or equal to 3, or whether you use LTV/CAC ratio to compare one investment opportunity to another, factoring in a company cost structure into the interpretation of that ratio is crucial to avoid ruling out false negatives or ruling in false positives (all the more if CAC Payback Period
*does*beat the benchmark).

# What’s the bottom line?

Keeping in mind some simple benchmarks and one-size-fit-all guidelines remains a reasonably valid strategy to monitor how the market is doing and do some reality/sanity checks at a high level.

However, as SaaS companies keep burgeoning, taking over all verticals and niches, with always-more specific business models, it is essential that we—founders and investors—refine our common understanding of Key Performance Indicators.

The sharper our ability to spot risks and opportunities, the greater our chances of building success at scale.

After all, metrics and benchmarks don’t make or break companies: we do.

There is a lot more where that came from and we will likely follow on with a second post to go one step further in the analysis of LTV.

In the meantime, if you find this article worthy of discovering and reading for other SaaS founders & investors, feel free to give it some claps 👏👏👏