In their last earnings conference call (Q1 2012), Angie's List (NASDAQ:ANGI) claimed that costs per acquisition are going down and renewal rates are going up. This sounds good at first glance, but the problem with terms like "acquisition" and "renewal" is that they are arbitrarily defined by management.
"CPA" is Meaningless
Although the cost of acquiring customers has gone down, this is only because the average price that ANGI charges customers has gone down, and in some areas, they are giving away their service for free. Yet ANGI claims victory in "acquiring" customers who do not pay for their subscriptions. Due to these ambiguities, and the shifting of ANGI's revenue sources from member subscriptions to service provider contracts, the most straightforward way to calculate the efficiency of management's spending is to derive the cost per dollar of revenue that results from their sales & marketing spending. We will get back to this later, but first we need to explore what "renewal" means.
"Renewal" is Meaningless
If you look at the fine print on their SEC filings, you'll see that their renewal rates exclude monthly memberships.
(5)First-year renewal rate reflects the percentage of paid memberships expiring in the reporting period after the first year of membership that are renewed, and average membership renewal rate reflects the percentage of all paid memberships expiring in the reporting period that are renewed. Renewal rates exclude monthly memberships. [Angie's List Form S-1/A Registration Statement]
Since monthly members almost never renew their subscriptions for more than a year, the real renewal rate is significantly lower than what is represented in their SEC filings. However, ANGI does include monthly members in their tallies of total paid memberships and of membership growth. It is disingenuous to exclude monthly members in some statistics and not others. Furthermore, their renewal rate is not the same thing as their customer retention rate. This is because ANGI is the recipient of a high number of refunds and chargebacks (a natural consequence of their high levels of customer dissatisfaction). These chargebacks cause membership loss prior to the "expiration period" and are thus not included in ANGI's definition of what a renewal is. For example, someone might purchase an annual subscription, which is set to expire the next year. After they realize they've been duped into purchasing a bad product, they demand a refund and/or file a chargeback with their credit card company. ANGI does not count this event as a failure to renew, thereby inflating their renewal rates.
So What's Really Happening?
At first glance, it may appear that ANGI does not provide data regarding how many members they lose per year, but they do. You would think that they would display it in the same data table that includes "Gross paid memberships added in period." But instead, ANGI relegates the "memberships lost" statistic to the fine print in an all too obvious attempt to hide unflattering data.
By adding these fine print numbers to our spreadsheet, we can get a rough idea of the real member retention rate. Since the "Total paid memberships" statistic refers to a measurement taken at the end of a given period, we can say that "Total paid memberships (End of 2009)" is equivalent to "Total paid memberships (Beginning of 2010)". So the equation for finding the minimum real membership retention rate is:
Click to enlarge.
When we apply this equation to our spreadsheet, the results are surprisingly consistent. For every audited year available (2008-2011), the minimum retention rate is (59±1)%. In other words, every year, the number of memberships lost during that year is equal to about 41% of the total memberships at the beginning of that year.
This is important because it shows that the rate of member retention in the absence of a sales and marketing budget will be somewhere between 59% and the distorted renewal rate reported by management (about 75%). Using these two numbers as book ends, we can model the cost per dollar of revenue using this equation:
Note that since we do not know the shape of the cost per dollar of revenue curve, we will assume for simplicity's sake that it is linear. But a more in-depth analysis would compare the economies of scale garnered by larger advertising budgets versus the diminishing marginal returns to advertisement saturation.
When applying the cost per dollar of revenue formula to our spreadsheet, we discover that ANGI is currently spending about $1.65 per dollar of revenue assuming an angry, rapidly decaying customer base, or $1.96 per dollar of revenue assuming a loyal, slowly decaying customer base. These figures only include sales & marketing expenses, so even if ANGI spent $1 per dollar of revenue, they would still be losing money. You may be wondering why sales & marketing spending becomes more efficient given an angrier customer base. This is because sales & marketing spending is a fixed variable and a quickly decaying customer base means that a greater proportion of membership maintenance and growth is due to effective sales & marketing spending. Whereas if the customer base is more loyal, maintenance and growth of membership levels is less attributable to sales & marketing spending, and is more attributable to natural retention rates.
Obviously at some point, ANGI must stop spending $2 in order to buy $1 of additional revenue. This strategy is unsustainable, as it would eventually lead to infinite losses. Assuming a linear equation, we find that if ANGI spent the minimum sales & marketing budgets necessary to maintain revenue at existing levels, their pre-tax net profit would be somewhere between -$20M per year and +$2M per year. This would suggest that their business model is anywhere from extremely unsustainable to barely sustainable.
In the best case scenario in which ANGI is capable of generating pre-tax profits of $2M per year, applying a generous 10x P/E ratio yields a valuation of $20M. Based on 56.96M shares outstanding, this would indicate that a fair price for ANGI is 35 cents per share, assuming a loyal customer base, a 10x P/E ratio, and stable future cost of revenue. However, the cost of revenue has been steadily rising, the customer base is very angry, and the qualitative aspects of this company don't justify a 10x P/E ratio, which is why the actual value of this company is probably well below 35 cents per share.
Disclosure: I am short ANGI.