Customer Acquisition Costs (CAC) Data and Trends.
Most ‘Business Modelling‘ fails because either (a) outlandish predictions for revenue growth are made or (b) little thought is given to Customer Acquisition Costs (CAC) thus creating very inflated margin expectations.
The CAC ultimately eats into the Gross Margin; if we end up with a thin Contribution Margin then the whole business model will be under duress.
Here, we cite some detailed CAC data from various different industries and layer in the revenue logic to justify each spend.
B2C CAC Data + Insights
- CAC Trends Up Over Time
- eCommerce/Retail Customer Acquisition Data
- Fintech Customer Acquisition Data
- Fashion Customer Acquisition Data
- Business Modelling – CAC Up or Down?
- More Customer Acquisition Cost (CAC) Posts
- Case Studies Cited
CAC Trends Up Over Time
Customer Acquisition costs generally go up over time. Early entrants in almost any market have a huge advantage to acquire customers before the rest of the industry wakes up!
But once a market becomes more ‘mainstream,’ competition increases and companies will be willing to bid-up for those customers through virtually any Paid Channel.
During the creation of a lot of projections for business models, mistakes are made on Customer Acquisition Costs (CAC) because the theory around it is:
- (a) hard to understand and
- (b) difficult to benchmark
Below, specific industry data is shared relative to each category, along with more advanced Business Model Analytics to try and share context of how the pieces fit together
In the interest of clarity, we classify ‘Customer Acquisition Cost’ as the ‘Demand Generation Expense’ – this means any expense related to acquire those customers, whether through Paid Ads or any staff who help acquire customers through organic means. It it not simply the costs of Paid Ads.
The caveat to all the below data is – despite it all being well-sourced – that doesn’t mean it was all calculated in the same way. Thus each data point, while informative, shouldn’t be taken as absolute.
eCommerce/Retail Customer Acquisition Data
The magic of the D2C (Direct-to-Consumer) boom has faded and changes in privacy standards/regulations are making it more restrictive to advertise on major digital platforms, both of which contribute to a rise of average CAC from $9 to $29 between 2013 and 2022.
That doesn’t mean that the retail/eCommerce business model is dead. Data from the same report shows that the average Repeat Purchase has gone from $28 to $39 in that same time period.
What that means is that the margin between the ARPU (Average Revenue Per User) and CAC (Customer Acquisition Cost) has shrunk in the short-term, but increased in the mid-term for Repeat Customers. It also means that Customer Retention is key in retail/eCommerce.
There will obviously be a great deal of variability in and around both the CAC and the Repeat Purchase #, as they are simply averages.
A lot of the Gross Margin calculation will depend on how much the good(s) sell for (Average Order Value) adjusted for the costs to fulfill the order (Shipping, etc.). The CAC ultimately eats into the Gross Margin, if we end up with a thin Contribution Margin then the whole business model will be under duress.
Fintech Customer Acquisition Data
Fintech is a beast when it comes to CAC.
First and foremost, anything to do with banking is historically relationship-based, which means that the Customer Relationship normally spans out over the long-term. As we can see in the graph below meaning that the average customer is worth A LOT to the Retail Banks ($300 – $600+ CAC).
Those same customers cost an order of magnitude less to the Challenger Banks ($30 – $50 CAC). And then there is Square’s Cash App, acquiring customers in its most recent quarter for a $10 CAC (and only $5 prior to 2020). Nubank’s CAC is the lowest of all the major Fintechs at this moment, a function of its its presence in Emerging Markets.
The discrepancy is seen relative to the business models of each and their ability to monetize those same customers over time:
- Retail Banks – make money, generally, on some combination of NIM (Net Interest Margin) and high fees across a variety of product categories (Banking Business Model Canvas). This means that over the lifetime of a customer, Retail Banks can earn many multiples of the $300 – $600 CAC expense
- Challenger Banks – brand themselves on being ‘non-banks’ with low fee models, which range from monthly subscriptions (Revolut) to more payments-based models (Chime), or a hybrid between models in the case of the Revolut Super App. This leads to some form of model that can become profitable on a per customer basis each year with CAC in the $30 – $50 range
- For Cash App, ARPU is estimated at $50 per user (Block [Square] Business Model Canvas). Furthermore, Gross Margin per customer is contingent on what products each customer uses within Cash App, meaning a strong profit can be made per customer at $10 per user
Thus, any judgments made about CAC in Fintech need to be classified against what the prospective business model is. The closer it is to Payments, the thinner the margins and the lower the expected CAC. The closer it is to Lending, the fatter the margins and the higher the expected CAC.
Gross Margins in Fintech involve more complex calculations related to the technology costs to onboard new customers and execute the service; there is a level of scale required to reach reasonable cost efficiencies in this model, which is why they invest so much in technology.
Once they achieve scale, however, certain customer Cohorts can become massive profit centers for Fintechs which is why the arms race between Retail Banks and Fintechs is so intense.
Fashion Customer Acquisition Data
While fashion overlaps with retail/eCommerce, the dynamics of the market are unique compared to most other markets.
The lower end of the market (Fast Fashion) operates on relatively low margins and high volume, while the higher end of the market (Luxury Fashion) has high margins and low volume. The sheer volume of fashion spend each year – as fashion is a necessity – means that new models are always arising in fashion to try and ‘disrupt’ the industry.
Luxury fashion marketplace Farfetch was a headline grabber for several years following its IPO (Q3 2018) thanks to a mix of both business model and technological innovation.
They acquired customers for ~$100 at scale and made a lot of money on the average customer over time due to their ability to retain the majority of their New Customer base.
Farfetch boomed in the pandemic years like most other digital platforms, reaching a market cap of $20B. Since Q1 2021, they have lost about 90% of their market cap (currently ~$2B) thanks in part to a mix of factors, among which is a substantial increase in their CAC, causing an erosion of their margins and ultimately profitability.
Farfetch uses LTV as a core business metric, with it being often-mentioned in their quarterly calls. When they IPO’d in 2018 it was disclosed as shown below. The Lifetime Value is based on ARPU (Average Revenue Per User) calculations. These ‘Unit Economics’ for Farfetch had been strong since their inception, but have come under pressure in recent quarters.
Farfetch has a Take Rate of ~30% off the top-line price of the good. ‘Revenue’ in this case is actually ‘Net Revenue,’ which is ‘net’ of fulfillment, payment fees and taxes because the consumer/retailer pays those costs. This is important to define because AOV (Average Order Value) averages around $600 over time.
But $600 is not ‘Revenue;’ the approximate Revenue from a $600 transaction for Farfetch would be $180 ($600 x 30% Take Rate).
Most models account for the fact that the average customer who is acquired will purchase more than once.
If for example, we projected 2X purchase per year, then Revenue equates to an ARPU (Average Revenue Per User) of $360 ($180 x 2), which is then projected into the future to calculate LTV (Lifetime Value).
Below is one such example of an investment model looking at the potential NPV (Net Present Value) with a CAC of $300. Even at these levels, there is a case to be made that a company can make money due to the loyalty/repeat purchase behavior – Farfetch had ~55% Repeat Customers at the time of IPO.
Farfetch’s Gross Margin has remained relatively stable around 44% in the last several years, showing that the above model may be a little bit rosy.
But Contribution Margins have decreased steadily from the mid 40% range in its early pre-IPO days down to 32.7% in Q1 ’22. This means that increase in CAC (combined with other factors) has bit into its margins materially enough to erode the overall profitability of their business model.
As a percentage of Digital Platform Services Revenue, Demand generation expense was 21.1% compared to 18.9% in second quarter 2021. The increase was driven by our investment in acquiring and engaging customers in paid channels, and reflects continued cost inflation in digital marketing channels….Farfetch Q2 ’22
The entire fashion sector – both online and offline, luxury and fast – has come under pressure in the last few years. The Farfetch CAC data shows that markets can change significantly over the long-term. Farfetch maintained a stable and respectable CAC around $100 for many years, and rose to prominence in the fashion world around that business model.
As the market has changed structurally and those costs rose, they have not been able to compensate with increased revenues enough to maintain their margins. Material changes in CAC can even make multi-billion dollar digital brands shed the lion’s share of their market cap within 12-18 months.
Business Modelling – CAC Up or Down?
When putting together a model that is relying on certain assumptions and trying to make projections into the future, the question is, should CAC be modelled up or down?
First and foremost, any model should attempt to benchmark CAC based on some kind of industry research or firsthand experience.
As stated earlier, CAC is defined as the ‘Demand Generation Expense’ related to acquiring customers. It is not just Ads or explicit expenses, it is also linked to the time spent by any staff to acquire or onboard customers.
That doesn’t mean that there are any standards for how CAC is measured or stated. There is a level of subjectivity to really any CAC number; nevertheless, an educated guess is better than a random one.
Secondly, while yes CAC costs tend to go up over time, they don’t always do so in a predictable or linear way. Furthermore, by the time a company gets 3 – 5 years into the future, they can afford those costs on the back of some sort of presumed profitability.
Thus, there are a few factors to consider in relation to this question:
Are there any organic levers of growth?
Organic levers could be:
- a hyper-engaged community that become Brand Ambassadors and drive new customers
- a really cool product that generates a lot of online buzz
- some kind of referral/network lever that creates hyper-growth without advertising
If the answer is yes, then there is a chance that the initial CAC will be much lower than what competitors may have. It is still NOT a good idea to benchmark CAC too far below an industry benchmark at the beginning – then expectations are set so high and the business model might be hemorrhaging cash relative to expectations.
But if there is some data to show that organic levers are driving growth at a low CAC multiple relative to the industry, there is no reason to model the CAC up into the future. Cash App is still acquiring its customers below the industry average with 45M MAUs (Monthly Active Users).
What is the projected LTV (Lifetime Value) multiple relative to CAC?
We saw above how Farfetch benchmarked CAC off of ARPU (Average Revenue Per User) to target a LTV (Lifetime Value) multiple of at least 3X (or 300%) Revenue earned through the lifetime of a customer versus CAC.
In the case of Farfetch’s CAC eventually increasing significantly, this doesn’t necessarily kill a business model, but we saw that it could. In the last several quarters, they have been under pressure and going forward investors will be waiting to see if they can turn it around.
Modelling CAC up over time can be a good stress test of any business model. If increased CAC growth in the future blows up a business model, then it needs to be tweaked.
What % of New Customers are projected to become Repeat Customers?
This is probably the most important question. If the % of New Customers who become Repeat Customers is high, then an aggressive CAC spend can be justified even with somewhat uncertain Unit Economics. If a large % of New Customers end up in Churn, then the business model will fall apart in the vast majority of cases.
We saw how in eCommerce, Average Repeat Purchases have actually gone up ~40% in dollar terms (ie. more revenue) over the same time period where CAC has gone up ~200% (ie. higher costs).
The firms that would have survived the best during this period would be those with the highest Contribution Margins, a factor that is driven by how much they outlay on CAC.
Therefore, when determining whether to model CAC up or down, look at strategies/costs for Customer Retention in tandem. Yes, Customer Retention costs could eat into margin, but needing to constantly acquire New Customers will eat into margins a lot more!
Scenario Analysis – Finding the Balance
Different business models will have very different CAC costs, as we have seen in this post. The idea is to look at different CAC scenarios to understand how it effects the business model over time. Business Modelling can help identify different structural advantages/disadvantages in the model, from which various strategies can be deployed to adapt.
Overall, we can see how fickle and difficult it is to predict CAC over time. We have also seen how it can make or break business models in the short, medium, and long-term. That’s why working with some form of relatively accurate CAC assumptions is so important when modelling a business, new or existing.