In several posts (Semrush (SEMR): Recasting Expenses As Intangible Assets For ROIC (2/3), for example), I have discussed companies that invest in growth mostly through intangible investment through the income statement (e.g., S&M, G&A, R&D) rather than capital expenditures, and I have highlighted how these investments are incorrectly categorized as expenses under GAAP. If adjustments for this income statement investment are not made, ROIC numbers can be wildly separated from economic reality. We want dependable ROIC estimates, so we need to make these adjustments.
But these intangible investments are also related to another topic near to my heart: valuation. When we are making intangible investment adjustments to correctly estimate profits (numerator) and investment (denominator) to get to ROIC, we are essentially asking: “does this business generate returns on its capital investments (intangibles included) that make me interested in owning this stock?” But why not go one step further and say: “based on the returns generated by these investments in S&M, G&A, and R&D, what is the business worth?
This is what Dan McCarthy‘s work focuses on: Customer Based Valuation. He describes this as a bottom-up rather than top-down approach to valuing a business. This makes a lot of sense because ultimately any company is simply an assemblage of customers, or more specifically, an aggregate of the profit streams from those customers, plus the capital used in the business to get and keep them. So why not focus on the customers for valuation? Forget revenue and cash flow growth and cap ex assumptions, which are often over-simplifications and wildly arbitrary. Try to estimate the aggregate value of individual customer economics – that’s what the business value actually is. I have veered into this area a bit in some posts where I discuss customer acquisition costs, customer lifetime value, customer retention, etc. But McCarthy is the expert on these topics, and below is a link to a CFA talk he did in 2022 that introduces his methodology and applies it to a few examples.
I highly recommend watching the video (it’s about 48 minutes), and checking out some of his papers and other presentations from the links I provided below. Here is a quick summary of my notes from this CFA talk:
CFA Society – Atlanta Series
Enhancing Valuation Analysis Through Customer-Based Corporate Valuation
The goal is to leverage consumer behavior insights to enhance our ability to predict revenue and to use models to better understand the unit economic health of businesses.
McCarthy specializes in these methods. He models customer adoption for corporation valuation. He has used this methodology as a consultant and also started a predictive analytics business, which was used on 250+ businesses. It was sold to Nike in 2019, which is still using this model. McCarthy is still consulting for private equity firms using this methodology to understand unit economics and for valuation purposes.
Customer Based Corporate Valuation (CBVC)
Any standard corporate valuation book – McKinsey, Damodaran – will say the value of the equity of a business has 3 parts: (1) operating assets, (2) Non-operating assets, and (3) Net debt
Customers drive all of this, yet investors typically don’t spend much time on customers. For example, they don’t look at customer transaction data. But the prevalence of enterprise/CRM software has made this much easier today, and this is what McCarthy focuses on, transaction data.
CBVC is a basic accounting identity. It breaks down revenue into (1) adoption/retention, (2) purchase frequency, and (3) customer spend.
If you estimate the flow of customers and retention of them (i.e., looking at historic transaction data) you can fairly accurately estimate the size of the customer base over time.
If you use the data to estimate purchase sizes, now you have revenue. Then make gross margin contribution assumptions.
Once we make a customer acquisition cost (CAC) assumption and margin assumptions, we have what we need to get to a basic NPV analysis by applying a range of discount rates, much like what we do for a regular DCF.
Need 4 things: (1) predictive model for flow of customers, (2) model for customer retention, (3) model for purchase frequency, (4) how much do customers spend, when, and how often.
We are basically taking a marketing model from the marketing department and bringing it to finance/valuation world. We are predicting revenue far more accurately now, which helps a lot with valuation.
Some definitions:
Customer Lifetime Value (CLV): The NPV of all lifetime profits from the customer. Includes an assumption about how long you have the customer for (retention), customer acquisition cost (below), and variable margin assumption. Obviously, the longer the customer is retained the higher the CLV (assuming customers are profitable). Try to keep the number of years for your retention period short, like say 3 years – it’s more conservative.
Customer Acquisition Costs (CAC): What it costs to acquire or keep the customer – usually taken from Sales & Marketing expense plus some portion of G&A and R&D. A simplified way to calculate this is to look at new customers acquired in a period divided by the spend of the above items in that period.
Payback Period: How long it takes for the company to get the CAC back in profits (not revenue). Shorter the better. The longer the payback period, the more speculative the entire exercise becomes. It should be a few years or less.
Post Acquisition Value (PAV): PAV is the value of the customer before CAC, usually in present value terms. If we compare the PAV to the CAC, we get a multiple, the higher it is, the higher the return* on the CAC. For example, 4x would be PAV of $400/ CAC $100. The company spends $100 to get a return on that $100 of $400. The higher the better. Also described as Return On Customer Spend:
This is basically a customer unit-level return on invested capital estimation.
McCarthy provides some examples with real companies:
Revolve Clothing
Farfetch
Some Things To Watch Out For:
CAC can change a lot over time – e.g., keyword costs, big traditional marketing investments like television, hiring a large sales team when online customer acquisition is petering out.
There are different approaches to calculating CAC, i.e., which expense items to include, and whether to include spend on existing rather than new customers. Companies will play games with these assumptions to try to show lower CAC and higher return rates/multiples.
Which cohorts are the data from? Can vary a lot through time and different customer cohorts.
Selective disclosure – companies will show the KPI data they like and that supports a good valuation case, and omit data they don’t like. Be aware of this and assume if not disclosed, it’s not good.
Lag issues – if there are delays between getting customers and revenue, or revenue and expenses, this can skew your assumptions, so be aware of this possibility and adjust for it.
Gross vs. Net acquisition costs – Some companies remove the initial customer acquisition cost from the ongoing cost (e.g., the cost of the bike is omitted by Peloton), must think about this.
Companies will do everything possible to show smallest CAC possible.
All of this is much easier for subscription businesses because of the recurring nature of revenue, customer retention, and unit economics. For non-subscription businesses, this framework is far less reliable.
For private companies there will be far more data available to do this analysis, this is why private equity firms like it.
This approach is especially useful for businesses that are not (GAAP) profitable today, because can help you gain confidence if/when they will be profitable.
Be conscious of Total Addressable Market – a business can have great unit economics but hit a wall because of a small TAM and stop growing, which means projecting existing unit economics into the future via growth is a big error.
The Bottom Line
I like this framework because it gets you focused on what matters: customers. Every time a company acquires or loses a customer, the value of the business changes. In fact, this idea can even be extended to: every time a business interacts with a customer its valuation changes. Customer economics will drive the capital productivity of the business in the long run, there is no way to get around that. If the unit-level economics of a business are good, e.g., high customer value vs cost to acquire and maintain them, the ROIC of the overall business will likely be good too. Businesses with higher capital productivity can grow profitability more easily, and therefore will have higher valuations. I think great stock investors understand the relationship between customer economics and valuation intuitively. While they may not create models like McCarthy, they get 80% of the way there with this intuition, and it’s a valuable skill to have when investing in stocks for the long-run.
Links
Dan McCarthy’s LinkedIn page
“A General Framework For Customer Lifetime Value” by Dan McCarthy (presentation) which includes definitions of the terms mentioned above
A list of Dan McCarthy’s academic papers
Here’s a specific LinkedIn post McCarthy did in which he applies his framework to Instacart using the S-1 filing for its recent IPO, which contains a lot of KPI information.
A Harvard Business Review article from 2020 by McCarthy called How to Value a Company by Analyzing Its Customers
McCarthy’s 2020 paper How To Value A Company By Analyzing Its Customers