Unpacking LTV and CAC for growing startups
For our first guest post we’re thrilled to welcome Rachel White who recently joined one of our accelerator Sprint Weeks to talk about SaaS metrics and how this can impact your business model. Rachel is a CFO for growing tech companies previously at Practice Ignition, Hotelclub.com and Sun Microsystems.
For founders LTV, Life Time Value, and CAC, Customer Acquisition Cost, are two of the most important financial metrics of your business but often not very well understood in the early days as you grow. They’re also not metrics governed by any accounting rules (they’re a little too new for that) and metrics or unit economics that your investors will want to know about it.
There’s many different ways to measure them, much of which depends on your business model.
One use case I've used them for is an “ROI” metric on the deployment of capital (effectively a proxy for profitability). The ratio of LTV to CAC tells you whether you’re investing for growth or focused on reaching cashflow positive (or in deep trouble — read on).
In Australia, Xero does the best job of reporting the underlying metrics as a public company, they’re a case study I use a lot.
So, how do you measure LTV and CAC?
The first step is to pick a definition and be consistent. You’re looking for trends over time, versus a static number at a point in time. How does it move? What’s driving that change? This principle is the basis for all non accounting metrics reported externally.
Financial literacy of board directors is a constant theme in non-executive director’s education. For tech investors, becoming literate on the nuances of these metrics is equally important.
Life Time Value
In simple terms, this is the total spend of a customer over the life time of that customer. It differs from financial metrics in that it’s not bound by a 12 month reporting period.
There’s a lot of variability in this at an individual customer level. You’re looking for the average over your whole customer cohort, or major subsets of it. That subset analysis is very powerful in making decisions on where to invest in growth. For that reason a good financial or data analyst should be part of your finance or revenue team’s capability.
There’s two elements used to estimate the life time value. You’ll never get it exact — remember, you’re looking for long term trends and what’s driving the change.
- (a) Average Revenue per month per customer (known as ARPU)
- (b) Number of months the customer stays with you.
- a * b = Life Time Value.
Sounds simple? Read on...
Average Revenue per month (ARPU)
For subscription business models, this is fairly straightforward. What was the average subscription amount that month for all active customers?
The main twist here is making sure you spread out the value of contracts (3 or 12 months) billed in advance. Systems like Chargebee and Recurly have taken leaps and bounds over the last few years here. One of the factors that allows you to escape spreadsheet hell in your metric reporting.
For usage models, you’re looking for the average spend per customer per month. I average it over 3 months to remove the variability month to month in the metric so the underlying trends emerge. This can be price per transaction or “clip the ticket”.
The other metric that’s helpful for these business models is the number of months / year a customer is active. It’s also a useful indicator of product “stickiness”.
Hopefully this point is self evident — all analysis excludes VAT / GST included on invoices. Yes, I have seen people try and call that revenue before!
Number of months active
So where’s your crystal ball, I hear you say, on how long a customer will stick around for? Good question.
For this one, there’s a convention used that looks at the monthly churn rate.
How many customers churned this month, as a percentage of the active customers the month before.
You had 100 customers churn this month, with 5,000 active last month. That’s a churn rate of 2%.
That means, on average, a customer is staying for 50 months. At 100 customers churning per month, it would take 50 months for 5,000 to churn. How this is done is to divide 100% by 2% to get to 50.
Xero’s churn rate (the lowest rate in B2B given accounting software is perceived as a “must have” if you’re an SME) was 1.1% this year. That means their customers stay with Xero for 100 / 1.1 = 91 months, or 7.5 years.
For subscription models, again this is straightforward. A customer is active any given month until they cancel. So it’s very easy to count number churned and number active customers. The primary variability here is whether you use churn in the month they cancelled, or when the subscription ran out — often that can be the month afterwards. I use the month the subscription ran out as it then compares to financially reported revenue in terms of the ARPU calculation. For operational teams, this can get confusing as they look at it in the month it cancelled — so within your operational management you may need both.
For usage models, I look at how many customers were active in a given period. For business models with steady activity throughout the year, I use 3 months. Where there’s a heavy seasonality factor, I use 12 months. This is where I’ll bring in a factor on how often a customer transacts over the 12 months to ensure they really are active.
On churn, how long is someone inactive before you consider them “churned” in this case.
Again, there’s a lot of judgement here.
Life Time Value: Gross Revenue vs Margin
Where you have very little in the way of direct costs against revenue, you don’t need a distinction between the two. In this analysis I’m excluding hosting and customer success team costs.
However, if you do have a substantial 3rd party cost base, then you need both.
Let’s say you use a 3rd party to deliver a core element of your product, versus your own code. You pay the third party a cost for use of their platform, that could be per transaction or percentage of revenue as a couple of examples.
In this I don’t include the subscription costs for your business management tools. That may include salesforce, hubspot, workflow management tools for customer success etc.
A good example of direct costs to include are Stripe fees if facilitating payments is a core part of your model.
Why does this matter?
Those costs are normally 30%+ of your revenue for that product line. So they’re a big factor on cashflow.
Life Time Value (gross revenue) tells you about top line growth. Movements in this are very helpful in determining how pricing changes and customer mix are moving the trend line.
Life Time Value (margin) is what you need to use to determine the health of your deployment of capital.
One of the judgement calls you need to make is whether you need both. For Xero, I don’t as they are predominantly subscription revenue based with a very low direct cost base. Over the last 10 years I’ve made this distinction twice for models where they did have a large direct cost base.
Another example would be the driver incentives paid by Uber. Based on the analysis done during their attempt at an IPO in 2019, this was the primary factor that derailed their first attempt, given the assessment was this made their model unprofitable.
CAC — Customer Acquisition Cost
For any business with a long tailed revenue stream, there’s a timing mismatch between revenue and costs.
Revenue comes in over time — one of the things that makes tech companies so valuable, as it’s reliable and predictable.
The upfront cost of acquiring customers is substantial. The scale up capital raise is often used to fund this mismatch.
Xero is now cashflow positive. The recurring subscription revenue from a large base funds the acquisition of new customers within the same reporting period. If they’d ever needed to get to cashflow positive faster in their growth phase due to external capital market factors, they would have dialed back on acquisition & product development to “balance” the bank account.
To calculate CAC, you need two elements
- (a) Amount spent in a time period on customer acquisition activities. This normally is a factor of sales and marketing spend
- (b) Number of customers acquired during a time period
- CAC = a divided by b
The traditional overheads allocation includes sales and marketing.
A portion of those costs will be spent on acquisition. The remainder will be on customer retention and company brand building.
For a company scaling up, the majority will be on acquisition. Over time, this percentage will drop as the other costs become more of a factor (mainly in the marketing area).
I average this over 3 months, as this deals with any “lag” type issues where money spent one month = acquisition the following month, for example. I may use a different time period if there’s seasonality factors involved. Again, this averaging process smooths out variability so the underlying trends emerge more readily.
The mistake I see people make here is to only include program costs, such as digital ads or trade shows miss the salary cost of people on staff who manage these programs – the biggest cost of all.
From an investor reporting perspective, these costs need to be included to give an accurate picture.
From an operational perspective, when you’re assessing CAC between channels, I use a standard labour rate, either per day or per hour. Again, it doesn’t have to be precise. What it does do is ensure your team are deployed to the most effective channels. Vs having the entire company turn up to a trade show (back when we had those)……..
The other upfront cost that needs to be considered is the onboarding process. For many B2B products, to ensure it’s “sticky” (ie: has low churn), there’s a large overhead in onboarding. This is separate from account management that’s done to ensure the customer stays sticky once they’re up and running.
This isn’t a standard metric used in the industry, so what you call it will be based on what you call the process within your business.
In a recent example where this was as factor, I reported both CAC and onboarding as separate items, that combined into a total upfront cost / customer.
Why was that relevant? Good onboarding was labour intensive but led to very low churn rates. To compare life time value and CAC in a meaningful way, it needed to be factored into both sides.
Number of customers acquired
For SME B2B, this is normally fairly straightforward, how many customers signed up this month. One question is do you use sign up date or subscription start date, for example. For the sales team this will be sign up date, where as for ARPU calculations (i.e. when the customer is considered active) it’ll be the subscription start date.
For enterprise customers, this gets a lot tricker as you sign up a few of them, well spread out. Here, I tend to look at number of lines of business signed up within an enterprise, after the initial engagement. While this is more of an onboarding versus acquisition cost, it is a significant upfront investment to get your product up and running within their environment.
Often in enterprise engagements, you can bill for the onboarding element – for the purpose of these calculations I treat that as a reduction in CAC, versus revenue, as it’s not recurring and thus isn’t part of your valuation multiples.
LTV / CAC Ratio
Once you’ve got through all of the above, the rest is a fairly simple ratio. You take LTV (margin) and divide it by CAC.
If your LTV (margin) is below 1 you’re in deep trouble! Your product will never make money and require never ending capital raises to fund the company without ever producing a cashflow positive result. Warren Buffet has been known to refer to these models as “ponzi schemes”.
As a guide, if your ratio is between 1 and 3, you’re investing for high growth — if you’re near 1, you need a plan to be getting towards 3 in the foreseeable future. It will take you longer to get to cashflow positive, however you will acquire more of your total addressable market (TAM) a lot faster.
If the ratio is between 3 and 5, you’re preserving capital to get to cashflow positive, usually within your current cap raising cycle. Or the external capital raising market has shifted and you need to preserve cash.
Let’s say you find a highly effective acquisition channel. Do you “double down” on it, thus this ratio drops? How that decision impacts this ratio is very helpful in making that decision. That needs to be done in conjunction with your external market analysis. However, it’s the closest I’ve come to developing an ROI type metric on deploying capital to support that decision rapidly.
In conclusion LTV, CAC and their ratio are imperfect tools but important signals on the health of your startup.
In one sense they can be used as a proxy on where you should deploy your capital effectively or at the very least inform you if your business is sustainable in its current model.