Open Nav Close Nav

Resources

Discussion on Unit Economics

By Lead Edge Capital Quarterly Letter Q2 2015

This post was originally included in our Quarterly Letter to our LPs in Q2 2015.

Deal Lingo – a Discussion on Unit Economics

Investors and entrepreneurs frequently throw out the term unit economics as a sort of ambiguous catch all term to measure the quality of a business. “Oh… the unit economics on that thing are amazing.” Frankly, based on some of the deals that are getting done out there, it’s clear to us that people either 1) don’t understand what unit economics are or 2) are taking leaps of faith on models that do not work today, but could maybe work in the future at scale and with enough investment. In light of this, we thought we’d share how we think about unit economics, and some of the quantitative and qualitative measurements we take when evaluating deals. We touched on unit economics briefly in our last letter but wanted to do a deeper dive, as its one of the key drivers in building conviction for our investments.

Three of the most important things we measure are the lifetime value of a customer (LTV), the cost of acquiring that customer (CAC) and the ratio between those two numbers, LTV/CAC. In this letter, we’ll be discussing these two metrics, how they relate to one another, and provide examples of how they work into our investment paradigm.

Defining the metrics

LTV or CLTV is defined as the total gross profit coming in from an average customer. This is not to be confused with lifetime revenue, which is one measure but does not tell the full story.  Naturally, we only get to keep what is left after we pay out any variable costs (COGS), which can be the cost of purchasing raw goods, customer support costs, or even server expenses we incur when we outsource to Amazon Web Services.

Obviously calculating LTV is a proxy because we never have a lifetime of data. What we do is track a group/cohort of customers over time and calculate what “average” behavior looks like. Of course, every cohort has some customers who fall off quickly and those who become loyal power users. Though we never have exact data, we can extrapolate by measuring how revenue churns and backing into lifetime value. A simple example: if annual revenue churns at 20% for a cohort of customers, then we know the LTV is 5 years.

CAC is defined as the sales and marketing dollars spent to acquire a customer. This includes both the money we spend on things like Google or Facebook, in addition to the staff associated with executing that spend. For software companies, it includes salesforce base compensation, sales commissions, and even the implementation costs required to get the software up and running.

For both metrics, LTV and CAC, it is important to be “intellectually honest” about calculating the fully load cost of acquiring a customer, and the true gross profit generated by a business. We often see mistakes in the methodology behind these figures, and therefore sanity check the definitions with entrepreneurs to ensure that we are defining these metrics in the same way.

Some of the best businesses in the world have high LTV/CAC ratios. The more lifetime value you get for acquiring a single customer, the more profits fall to the bottom line. Expenses like R&D and G&A tend to stabilize and trend downward as a % of total revenue over time, but it’s the S&M costs that can scale variably and reduce the profitability of a business, especially if you have low LTV/CAC ratios.

We’ve discussed the leaky bucket problem in the past, but when you have low LTV/CAC, it means that the customers you pay to get in the door don’t stick around for very long. This creates a leaky bucket on the revenue front, and requires the company to keep filling the leaky bucket just so revenues do not shrink. As a result, high S&M costs have to be sustained. Some of the best models require strong marketing to get going, but get the benefit of virality to decrease customer acquisition (re: Craigslist, Zillow, Uber, Airbnb etc.)

In summary, we look to find businesses with as high a ratio as possible:

LTV CAC Image

To put in the simplest terms, frequency drives LTV.  The higher the order frequency the higher the LTV.  Even better is when the company doesn’t have to spend any more money on Sales & Marketing to re-acquire the customer.  Think Amazon, Uber, Seamless or OpenTable… each time you transact with the company, you go directly to the app or the website. These are the best businesses since they have zero re-acquisition costs.

One thing that we always look for when analyzing repeat behavior is discounting. For a number of e-commerce businesses recently, we have seen discounting in the form of “coupons,” which gets buried in the sales & marketing expense line. Companies will often say, “60% of my sales in a given month are from repeat customers.” This is awesome except, when we find that most of those repeat customers are coming in through coupons where the company loses money on the transaction.  All this means is that the company is selling dollar bills for 90 cents, which we all know is not a sustainable business over the long term! It is compounded by the fact that this sort of couponing can attract the wrong type of customer, which may not have the same LTV profile of a customer who is willing to pay full price on the first go around. These companies are creating fake LTV and burying the reduction in LTV as an increase in CAC. This practice really distorts genuine customer behavior, and we have been very sensitive to these practices.

We are trying to find companies that have very quick paybacks (within 6-12 months) on the initial CAC and then have high repeat behavior with no discounting.  These are the businesses that generate the most profits over long periods of time.  Examples of businesses like this are Uber, Amazon, Alibaba, Seamless/Grubhub, BlaBlaCar, OpenTable, Catawiki, eBay, Stubhub and Facebook.

In our opinion, there isn’t a golden ratio between LTV and CAC that gives a green light for one deal and a red light for another. The relationship between these two numbers can also change over time, for better or worse, so it always comes back to finding businesses with strong, defensible LTVs with low CACs and high customer repeat rates because the offering/product is so compelling.

The lifetime of lifetime value

It’s important to highlight that not all lifetime value is created equal. The defensibility of a company’s lifetime value can evolve overtime, and therefore when evaluating an investment, we not only think about the shape of the unit economics today, but whether those unit economics can persist over time. We also like to build in some “margin of safety” in our assumptions, so that in the case we are wrong, our unit economics can still hold up.

For a novel but simple software offering, the LTV of a customer and associated CAC might be pretty compelling at first, but overtime, the software gets commoditized, and the LTV goes down due to pricing pressure and CACs go up due to new entrants in the market.

On the flipside, a business like Workday, which develops and sells ERP software, has high lifetime values because its product is highly complex and its implementation is resource intensive: customers can end up sticking around for 10+ years! Although, Workday’s CACs can be high given a long sale cycle and high switching costs from older ERP systems, it is clear the unit economics of Workday work well, as reflected in its $880M of revenue, 64% YoY growth and $100M+ of free cash flow for FY 2015. Though Workday does not have technically have the network effects of an internet company, the complexity of its software makes it very challenging for new entrants to create a viable competitor. This makes it easier for investors to believe in the sustainability of its LTV and CAC story.

The fall and rise of CACs

The second piece of our unit economic puzzle is understanding the lifecycle of customer acquisition costs.  The CAC evolution story is as follows:

The chart below depicts the story above and overlays the relationship between CAC and customer adoption. The red line tracks the cost of acquiring customers as we move through the customer adoption curve, while the area under the blue represents customer adoption over time.

CAC Chart

What can be dangerous about investing in growth-stage businesses is the bottoming out of CAC some companies see in the early adopter phase. The LTVs look strong because early adopters tend to be higher value users in general, and the CACs look compelling because those same early adopters need less convincing to try something. Naturally, most growth stage businesses we look at are “early” relative to their overall lifecycle, which brings us to our next point.

Building in margin of safety

They key is to build in margin of safety around unit economic assumptions, so that even if the LTVs go down or CACs go up, you are still in the clear on profitable customer acquisition.

We recently went through this exercise with a new investment. When we dug into the data, we saw that the company was generating enough LTV in its first three months to recover the full cost of acquiring a customer. We rarely see that sort of payback in the deals we look at, and we know from our benchmarks that many great internet companies have CAC payback in the one year range. The result was a very high LTV/CAC ratio.

The fact that we had this margin of safety in the unit economics was a key attribute in building conviction around the opportunity. What we typically find is that these sorts of specific indicators tend to be reflected in the overall capital efficiency of a business.

An example of a business with amazing CAC dynamics is TripAdvisor. Businesses built on user generated content are amazing because the reviews (what is being “sold”) are free to create. TripAdvisor does not need to solicit customers—people seek out the reviews themselves. Lastly, as the company has more reviews, it becomes nearly impossible to replicate, creating a moat around its LTV or the ad revenue it generates for routing eyeballs.

Tying it all together

From this thought piece, we hope you have a better perspective on how we measure companies. The truth is that given the rate of innovation for technology companies, it is nearly impossible to predict how all these variables play out, and we have really have no clue what is going to happen five years in the future for any of our companies. What matters is building a framework that enables us to distinguish what separates good businesses from great businesses, and building in enough margin of safety so that if we are wrong, our investments don’t turn upside down.

Even more importantly, it is about backing entrepreneurs who use data to run their business and think critically about measuring their own businesses the way we do. They should know these numbers like the back of their hand, and should build them into their KPIs and dashboards. When we see a strong grasp of these metrics and a constant tinkering of how to improve them, it gives us confidence that our management teams are thinking about value creation in the right way. These qualities are something we look for and test by having in depth conversations and getting into the weeds. The best entrepreneurs understand their businesses strategically at 50,000 feet and tactically at ground level, and they aspire to operate like their high performing peers with best in class benchmarks.

I am confident that by backing these types of entrepreneurs, and applying a repeatable, intellectually honest framework for evaluating companies, we can generate superior risk-adjusted returns for our investors. In the current environment we are in, it can be hard to bid against investors who are willing to build in less margin of safety and take bolder (read: crazier) leaps of faith than we are. As long as we stick to the playbook and are patient, I am looking forward to what is ahead.

Related Resources