Customer acquisition cost economics 101: valuing CAC improvement

Investors, consultants, and executives are all waking up to the paramount role of customer behavior in overall company valuation, following the lead of investor, consultant, and executive pioneers like Michael Mauboussin (especially this ), Rob Markey , and Tren Griffin, respectively. And while Peter Fader and I popularized customer-based corporate valuation (CBCV) through the academic work we've done together, through Theta (the company we co-founded), and through subsequent work I've conducted with Elliot Oblander, we too are simply following the lead of others (most notably Don Lehmann and Sunil Gupta).

What is interesting is that within the academic CBCV literature, most of the emphasis (i.e., the fancy modeling) has historically been on the behavioral outputs -- the speed with which customers are acquired, how long they stay, how frequently they purchase, etc -- while far less attention has been spent on what drove those outputs, such as customer acquisition spending and marketing earmarked towards repeat orders. Heck, most of the time, CAC is assumed to be constant, growing at a linear rate, or something else equally simplistic! What is interesting is that when we move outside of the CBCV literature and into the more mainstream marketing literature, there is a lot more literature on the drivers, but very little that ties those drivers to overall valuation.

This is not a gap I'm going to fill in this little note, but I did want to provide a bit of color on one such driver -- customer acquisition cost or CAC -- as I'm preparing the materials for my course this Fall (you could probably guess the title without my mentioning its name, but yes, it is called Customer Lifetime Valuation!!!).

What I'll do here is peel back the onion a little bit on some empirical truths about how CAC improvement maps to valuation that you may not have appreciated before this note, but that hopefully will seem obvious after the fact. I'll spend the rest of the note explaining why, and even provide you with a valuation model (spoiler: here it is!) for full transparency and to allow you to play around with the assumptions should you so desire.

Cutting to the chase, these are the main things I want you to take away from this:

1. CAC improvement is non-linear in its impact on acquisitions. (Remember, you're making the denominator smaller)
2. The impact of CAC enhancement on revenue growth is cumulative over time, creating incremental revenue growth that accelerates over time.
3. CAC enhancement matters more when a company is earlier in its lifecycle than later.
4. More fundamentally, I hope this gives you a view of how you can think about the impact of actions on valuation, whether they be CAC improvements or something else. Calibrate models like these to your data (or get someone else to!), so you can more deeply understand ROI / value creation.

Next, I'll provide some upfront context to understand how I carried out the analysis that led to these points before diving into these points.

The context

Before I dive into the results I'll provide some context.

First, what is CAC?

The definition that I will use for CAC is as follows: the total amount spent on customer acquisition in a particular period, divided by the number of customers that were acquired in that same period. While there are better definitions for CAC, this is a very common one that is particularly useful in the context of CBCV, because it allows us to easily move from CAC to a line item on the income statement in a standard valuation model.

The way I'll go about framing the impact of a CAC improvement is through a counterfactual analysis.

• First, I introduce a baseline -- what we would expect for acquisition, retention, ordering, spending, and all the other valuation drivers that get us to cash flows and fair valuation, assuming that we didn't see an improvement in CAC.
• Next is the counterfactual -- what we would expect for those same behaviors assuming that we were able to improve CAC.
• By comparing the baseline to the counterfactual along each of the customer-driven dimensions and how they roll up into period-by-period financials (and ultimately valuation), we can get a very clear view of the impact of the CAC improvement.

A few comments about how I go about this:

• I assume we're doing this for a subscription-based firm. The same principles apply to non-subscription firms, but it makes the modeling easier.
• I use exactly the same sort of modeling methodology that Pete and I used in our original paper on CBCV for subscription firms...
• ... with a few simplifying assumptions -- basically that everything happens at the end of each quarter (acquisitions, churn opportunities, revenue, etc). Again, the takeaways are the same when you loosen this, but the spreadsheet gets significantly bigger.
• To keep things simple, I also assume no non-operating assets and no cash flow adjustments (e.g., non-financial working capital, etc). Again, those refinements flow through while leaving the story told here largely unaffected.
• I assume the baseline CAC is static. In truth, CAC tends to increase over time, but again, this doesn't change the story.
• I assume the improvement in CAC is "budget neutral" -- in English, that the firm spends the same amount on marketing in the baseline and counterfactual worlds, so whatever money is saved through lowering CAC is plowed back into acquiring more customers. The net effect of the CAC improvement is then more revenue growth with the same amount of new customer marketing expense.

That is basically it. You can control everything -- the base case assumptions for the various processes, the nature (and even timing) of the CAC improvement, and the other various valuation drivers -- through the "Assumptions" box on the "Modified Valuation" tab of the spreadsheet , reproduced here (these are the sort of figures you may see from a B2B SaaS firm - high CAC, but long customer lifetimes, high ARPU and nice margins):

CAC improvement has a non-linear effect on acquisitions

This is going to sound very simple because it is. But remember:

[Gross customers acquired] = [Marketing spend] / [CAC]

When we improve CAC, then, we're making the denominator smaller, which is the source of the non-linearity I'm referring to. While a budget-neutral 10% improvement in CAC means 11% more acquired customers each period, a 30% improvement means 43% more.

The upshot of this is that if you have a firm that has been extremely sloppy in terms of how they have gone about their marketing historically, and you now have the ability to very meaningfully improve the efficiency of their marketing spend, the new customers you're able to acquire as a result may be a bit bigger than you might have expected at first blush. If you can improve CAC by 30%, that doesn't mean 30% more customers each period. It means 43% (=1/.7-1) more.

CAC improvement's revenue impact is cumulative

Imagine that you just acquired the aforementioned B2B SaaS company. Assume the firm is 10 years old that had done an atrocious job of managing their marketing spend historically. You invest in an experimentation platform and the incrementality tests you run suggest that you can reduce CAC by 30% going forward. You naturally want to know what effect this will have on the company's overall revenue.

It won't be 43%! Remember, lowering CAC means more newly acquired customers, but you have a (hopefully large) base of existing customers that are also placing orders that won't be affected at all by this CAC improvement. Over time though, all those incremental new customers that you wouldn't have acquired at that baseline higher CAC level add up.

Your gross acquisitions may look something like this (remember, we're assuming the company is now 10 years (or 40 quarters) old):

But your active customer count will look something more like this:

Despite lowering CAC by 30% and increasing adds by 43%, the first couple of quarters will show "only" 7% and 12% incremental active customer (and thus revenue) growth, respectively...

... but as time passes, the impact of that CAC improvement compounds. The customers you acquire today will provide you with revenue years into the future. Because of this, the incremental revenue from a CAC improvement a year from now is driven not only by the new customers you acquire then, but also the new customers you had acquired over the preceding year who are still with you today. This accelerates the revenue growth impact over time, and is why we see incremental active customer growth moving up from 7-12% in the first couple of quarters after the event to above 30% (and ultimately asymptoting to 43% over a longer span of time).

Executives or acquirers modeling the overall company-level impact of CAC efficiencies will want to keep this in mind. You really cannot just look at improvement within the first 12 months post-acquisition/post-change and assume that that is the improvement you should expect thereafter. The impact often starts small but grows over time. This point may seem simple but it is overlooked in practice surprisingly often.

CAC enhancement matters earlier in a company's lifecycle

As you would expect, the overall valuation impact that a CAC improvement has on the value of a firm will matter more, all else equal, the younger a firm is. If Amazon were to improve CAC in its US retail business by 50% today onwards, the effect on Amazon's valuation would likely be exceedingly small. Amazon has acquired much of the market that it was looking to acquire here in the US already. Retention and development of those existing customers is what will largely determine the valuation of Amazon's US retail business. There just aren't that many more new customers to acquire, so the incremental adds that Amazon would get would be small relative to large existing customer base.

The valuation impact depends on where your company is in its lifecycle. Before you go off spending top dollar on a corporate initiative to improve CAC, it could pay to run scenarios that assess just how much value such an initiative will create for your business, given where it is right now.

You can easily mess around with this through the spreadsheet , by modifying when "now" is (e.g., when you acquired the business and made the operational change). Simply change the "Quarter # when valuation occurs" assumption (which further assumes that this is when the CAC improvement occurs).

Let's consider a few hypothetical scenarios:

• If we acquired the company 10 years after commercial operations began (again, lowering CAC by 35%), the overall valuation impact would be 35% (cell C43).
• If we founded the company and improved CAC at the very beginning of commercial operations, the valuation impact would be 44%.
• If instead, we acquired the company 25 years after commercial operations began, the valuation impact would be a much smaller 21%.

The baseline and counterfactual active customer count evolutions associated with that third scenario are shown here:

Again, this is closer to the aforementioned Amazon US retail scenario.

Closing comments...

My hope is that this got you thinking a bit more critically about CAC, what it means for valuation, and how its impact on valuation depends on where the business is at the time of the change.

When you're thinking about your next very big change, I would encourage you to calibrate (and then run sensitivity analysis on) models like this one to more deeply understand what you should expect from such a change. And loosen those assumptions! Bake in the cost of the initiative that you expect to drive the CAC reduction. Change the goodness of the customers that are acquired if you are worried about those incremental customers being of lower value. Cap the number of years that you expect the CAC enhancements to last. This is all eminently doable and will likely lead to more valid inferences (and more buy-in from the CFO).

I really hope that more people start thinking about ROI this way, and allocating resources accordingly.