Is LTV to CAC the Nickelback of SaaS Metrics?
Is LTV to CAC the Nickelback of SaaS Metrics?
Why a metric that gets so much airtime isn't actually loved
LTV to CAC and Nickelback are both:
Overplayed
Unrealistic
Prone to error
Unchanging with the times
General, and devoid of context
Canadian
Today I’ll explain why LTV to CAC is flawed, yet a consistent recipient of misplaced love.
TL;DR:
It’s too compound of a metric for decision making purposes
It’s theoretical, and often unrealistic
People make some very common mistakes
Forgetting to Gross Margin Adjust your LTV
Using the wrong churn rate
Refusing to segment
The difficulty of landing a customer changes over time
General benchmarks are like general life advice
Bigger isn’t always better
What I’d use instead
CAC Payback Period
Net Dollar Retention
I’ve had a bone to pick with LTV to CAC for too long. And today I woke up and chose violence.
It’s too compound of a metric
If I wake up one morning, throw Silver Side Up into the tape player, check on my LTV to CAC, and want to change it, who do I talk to? Do I reach out to the head of sales who owns new ARR? Do I reach out to the head of customer success who owns churn? Do I reach out to the head of marketing who is spending all our damn money on customer acquisition? There’s no clear owner.
And when you can’t point out who owns a metric, it just becomes an amorphous blob, dancing through decks without a care in the world.
Compound metrics are not behavior changing. They may be helpful for a binary investment decision - a “go or no go” type of measurement. But for all the people trying to make the trains run on time, you need a specific person who has specific levers at their disposal to change underlying behavior. If not, you should put the metric in the corner.
Note: I highly suggest you read piece he wrote with on what makes a good metric:
LTV to CAC is theoretical, and often unrealistic
If you make the numbers dance, you can get your LTV to CAC to infinity. And with the exception of the Dutch East India Company, startups don’t live forever.
The only two things certain in life are death and taxes.
I always like to say:
“I pray for the confidence of a seed stage founder with eight months of sales history who quotes a nine year LTV to CAC.”
There are a million reasons startups fail. But one is getting over your skis on customer acquisition cost and running out of gas.
If you spend into the ground to acquire customers, you won’t survive long enough to see the ninth year you calculated on paper.
Let's get this circus on down the road
We're taking bets how far a tank of gas will go-San Quentin, by Nickelback
People make a few very common mistakes
Forgetting to Gross Margin Adjust your LTV
The biggest mistake is not gross margin adjusting what you expect to get from a customer. Remember - customers may cost less after you land them, but they still cost SOMETHING to keep around. And in SaaS, that’s around 1/5th (or 20%) of the revenue you get from them each year. Not insignificant.
In the formula above, we’re taking the ARPA (average revenue per account), dividing by the average account churn rate, and then multiplying by gross margin. This last step is crucial to get an accurate read.
Using the Wrong Churn Rate
Another mistake is using the wrong churn rate. As stated above, you should be using an account based churn rate, not a dollar based one. This is important because you might have a much better dollar denominated churn (or retention) rate, as higher paying customers subsidize the pots and pans who churn out more frequently.
Some metric wonks will push back on the “one man, one vote” idea and argue that you should use a dollar denominated gross retention figure (ending ARR / starting ARR from the same customers, without expansion dollars).
I’d contend that if you are also slicing customers by segment (SMB, Midmarket, Enterprise) you get around the majority of the abnormalities that could be caused by not using dollar denominated. Plus, I’d rather a stricter view of my churn if I am using the signals to operate the business - give it to me straight doc; and by segment, so I actually know where to look!
Refusing to segment
And a third mistake is forgetting to segment at all. This should be done by both region and customer type.
Procore commits an LTV to CAC crime here - while they do go to the regional level (getting themselves out of a felony), they don’t go down to the customer segment level (and deserve at least a parking ticket). This means that you have US based customers with 10 employees who spend $2,000 a year on the platform grouped in with US customers with 10,000 employees who spending $2,000,000 a year. The output is too cloudy, and dominated by the attributes of your bigger customers.
The difficulty of landing a customer changes over time
This is really important for companies who rely heavily on ad spend. You can’t blindly plow dollars into the same marketing source and not expect the well to eventually run dry.
CAC for the first 100 users is different from the next 1,000 users. If you put dollars into a campaign for a specific audience set, in the beginning, you’ll enjoy better returns on your ad spend. But, this result does not extrapolate to bigger budgets.
Meaning, “As you scale your ads, your return on ad spend generally goes down to a point where it won’t be profitable past a certain point.”
Various factors are in play here, for example, saturation of audience and how ad platforms usually find the lowest cost users first. So, it doesn't make sense to run a small test, calculate a LTV/CAC, and assume that will apply for a larger spend.”
-Source: Growth Eng Blog
At a more macro level, customers also experience diminishing Sales and Marketing returns when they try to “Cross the Chasm”. After you burn through the Innovators and Early Adopters, each new customer becomes incrementally harder to land. And then the sledding gets tough again once you’ve sufficiently saturated a market and are battling to win over laggards or displace existing competition.
In that way, LTV to CAC is more fluid, and more like water, rather than a static metric you can quote and forecast off of. It changes many times over a company’s seasons.
General benchmarks are like general life advice - lazy and not that helpful
Whenever you ask what a “good” LTV to CAC looks like, you’ll often get quoted an unhelpful 3 to 1 ratio.
But not every industry, or even subsegment of the same industry, has the same customer behavior (or rate of survival).
In B2C print media, 2 is pretty good.
In life insurance, 10 is normal.
In SMB SaaS, 3 is solid.
In Enterprise SaaS, 5 is expected.
Without any more context, you may as well just tell someone to not churn their customer. That would be more directionally accurate.
It takes too long to get an accurate read
Even if you do all the things above right - you gross margin adjust, you mind your churn rates, you segment… you still need to wait long enough to have sufficient data.
This probably means three or more years of selling history. Yup…+36 months!
Why this long? You need to give the cohorts you are measuring ample time to potentially churn. And if you have annual contracts that go out three years, it would be disingenuous to quote an LTV to CAC on customers who haven’t gotten the chance to make a renewal decision.
Yes, they technically will stick around by the letter of the contract, making their LTV at least three years. But even this is not indicative of true customer behavior (and def not satisfaction) and what they might do in the future, which impacts the future cash flows your business is ultimately valued upon.
In summary, companies make two potential foot faults here:
Imputing an LTV to CAC based on too small of an “n” sample size, and extrapolating it out over the entire customer base.
And calculating LTV to CAC on a cohort of customers that literally can’t churn.
As a rule of thumb, I’m skeptical of any LTV to CAC that quotes a number longer than the time the company has been selling software.
How the hell'd we wind up like this?
And why weren't we able
To see the signs that we missed?-Someday by Nickelback
Bigger isn’t always better
A massively good looking LTV to CAC could indicate you’re actually leaving growth on the table by not investing more in your sales and marketing machine to go out and acquire more customers. Since valuation is often tied to growth, you may be restraining shareholder value.
And from a competitive standpoint, you might be making life too easy for your competitors by not more aggressively chasing net new customers.
As a rule of thumb, if your LTV to CAC sky rockets to double digits, you may want to think about cranking up the investments in your go to market engine.
In fact, many VCs will cap your LTV to CAC at 10, regardless of what you give them.
My friends on the world wide web agree:
I asked people which metric played music with Chad Kroeger:
LTV, unless all years are prepaid in full, gives a false sense of security that can result in over spending on everything, especially expensive sales teams. CAC calcs are often full of lies resulting in a CAC that’s likely viewed with rose colored glasses on. CAC payback is a great but only if CAC is honest.
Compound metrics are a zero-interest rate phenomenon
Definitely LTV:CAC. It’s pretty much a meaningless metric unless you are a very mature organic. LTV as a whole is very theoretical.
LTV/CAC, especially for early stage companies. LTV ends up being a theoretical number.
What should you use instead?
I’d suggest relying on a combination of Net Dollar Retention and CAC Payback Period.
CAC Payback Period will tell you how long it takes to get your money back - very important so you don’t spend into the ground and not survive to see your estimated LTV to CAC horizon (poor cash management is a sure fire way to die)
Net Dollar Retention will tell you how much your customers expand each year over their lifetime, which in many ways, is an annualized, more timely version of LTV.
Using these two metrics together will allow you to triangulate how valuable a customer is on an annual basis and square that with the amount of time it takes to make that money back.
Nickelback has sold over 50 million albums. And LTV to CAC has graced at least 50 million board decks. But that doesn’t mean you should keep pressing play on either.
[Editor’s note: CJ actually loves Nickelback, but don’t tell anyone.]
A word from our sponsor Brex
The controllers I talk to are making big changes in the new year. They want to shift the perception of their role away from controlling and toward empowering and growing the business. It’s an uphill battle — the job title literally means “person who controls” — yet the best controllers I know are now CFOs because they could prove their value beyond cost control.
The current accountant shortage presents a golden opportunity. By streamlining processes, empowering teams to spend smartly, and maximizing efficiency, A-plus controllers can drive impact across the business.
Learn how in the controller’s guide to overcoming the accountant shortage from Brex.
Run the Numbers
Me and James Currier of NFX did a deep dive on Marketplaces and Network Effects.
Quote I’ve Been Pondering
“These five words in my head
Scream, "Are we havin' fun yet?"
-How You Remind Me, by Nickelback
What I’ve Been Reading
The VC Corner is a weekly venture capital newsletter, expertly crafted for investors, entrepreneurs, and startup enthusiasts.
As an investor at Mundi Ventures, Ruben brings firsthand insights from the venture capital world, curating content that resonates with over +2,000 Venture Capitalists, Startup Founders, and Business Innovators like you.
Subscribe to delve into the latest startup trends, venture capital news, and comprehensive analysis.
CJ- we at Procore slice/dice LTV/CAC by segment/stakeholder(buyer) and geography. We just disclosed the by geography one at the 2022 Investor Day. Difference between measuring how we operate vs what we disclose. Hope that helps.
One of the greatest (LTV/CAC) articles ever written (on Substack)