Angie's List (ANGI) has been a hotly debated stock in the investment community, with short sellers circling around. Many arguments have been dedicated to issues surrounding its business issues. Fewer people, however, have attempted to address the core of the long thesis on Angie's.
Henry Ellenbogen who manages the flagship fund at T.Rowe Price pitches Angie's List as his top idea and believes the Company has a superior business model and is poised for immense profitability after a hyper-growth phase passes. The core of his thesis rests on the immense profitability he sees in the most mature markets ran by Angie's. It is less clear to me whether Mr. Ellenbogen analyzed the assumptions used in Angie's cohort analysis in deriving the purported profitability. I believe Angie's has been misleading investors in its cohort analysis by misallocating marketing expense using demographic data. Its analysis overstates the profitability of its businesses in the most mature markets by understating, or underallocating, marketing expenses to those markets. At a market capitalization of over $1.6 billion, Angie's List is an incredibly attractive short selling opportunity due to the misunderstanding of cohort profitability by some of the biggest mutual fund investors in the world.
Angie's Hidden Assumption and Its Flaws
Hidden in the fine print, Angie's List discloses how it allocates marketing expenses to each individual cohort.
"Average marketing expense per market is calculated first by allocating marketing expense to each cohort based on the percentage of our total target demographic for all markets in such cohort, as determined by third-party data, and then dividing the allocated cohort marketing expense by the number of markets in the cohort at period end. We calculate this average per market to facilitate comparisons among cohorts, but it is not intended to represent typical characteristics of actual markets within the cohort. According to a March 2013 demographic study by Merkle Inc. that we commissioned, there were approximately 31 million households in the United States in our target demographic, which consists of homeowners aged 35 to 64 with an annual household income of at least $75,000. Approximately 27 million of these households were in our markets. The average number of households per market in our demographic target was 425,000, 435,000, 60,000 and 20,000 for the pre-2003, 2003-2007, 2008-2010 and post-2010 cohorts, respectively."
On the surface, it makes perfect sense. Angie's List has a national advertising budget and hence it should allocate its expense equally to each household it reaches in its targeted market. However, the analysts who modeled Angie's cohort profitability using the management's assumption failed to ask a critical question: is the lifetime value of a user in a 20,000 household market the same as the lifetime value of a user in a 425,000 household market?
Angie's essentially operates a local review and advertising business. It charges consumers a fee to find a reliable local service provider such as a plumber or electrician and it charges those same service providers to advertise on its platform. It is a valuable platform but the value of its service is substantially different in a small town versus a big cosmopolitan city. In a small city with 20,000 household, residents tend to know each other fairly well and the community is very small. There are typically very few plumbers and electricians serving the entire market and the likelihood of a rip-off is fairly low. By comparison, in a big metropolitan city like New York or Los Angeles, people probably do not have a plumber who lives next door and may not know a reliable plumber personally. It makes sense for consumers to pay in search for a reliable service provider and for service providers to pay to gain leads from companies such as Angie's List.
Naturally, the value of local advertising in a big metropolitan market is exponentially higher than in a small town with 20,000 households. For each individual user attracted onto Angie's platform through advertising, the lifetime value of a small town user is significantly less than the lifetime value of a big city user both from a subscription fee perspective and advertising revenue per user perspective. This is a demonstrated fact in Angie's financial statement.
Per Angie's disclosure, the average revenue per user in markets with 400,000 plus household is well over $125 per user while the average revenue per user in the smaller markets is in the $30 to $40 range. It is not because Angie's List only recently established business in those markets; it is because the value of local advertising in small towns is simply significantly lower. Neither users nor service providers will ever be willing to pay anywhere close to the level Angie's is charging in the big metropolitan markets that belong to its earlier cohorts.
Given that both data and common sense point to a fact that value of user in small markets is significantly lower than value of user in big metropolitan markets, Angie's accounting practice of allocating advertising expense based on demographic data becomes incredibly curious. Angie's is implicitly treating every household in its total addressable market the same by allocating its marketing expense demographically. Under the current practice, Angie's allocates $83 per user in marketing expense regardless of which market the user belongs. For users in the smaller markets with inherently less value, Angie's is still applying an $83 advertising spending to each and every one of them despite they may never be able to generate anywhere close to that number in revenue per user terms. This flies directly in the face of the accounting tenet, matching principle. Expense should always be matched against revenue. In allocating its marketing expense using demographic data instead of revenue numbers from each cohort, Angie's management overallocates marketing expense to the smaller markets and thereby underallocates marketing expense to its most mature and largest market cohorts knowing fully that the big mutual fund managers are watching the trend and profitability in those big mature cohorts. Bulls argue that the mature markets are incredibly profitable but they failed to see those markets cannot stand on their own. Their profitability is in fact subsidized by the losses from the younger cohorts. Independently analyzing profitability from pre-2003 cohort is inherently flawed.
Further, it is worth pointing out such expense allocation provides substantial room for management abuse. To illustrate, two extreme examples are presented here to show how management has the ability to significantly skew profitability of the pre-2003 cohorts simply by adding new markets.
Angie's can decide to add a million new markets each with 5,000 household. Let's call it Cohort Angie. The value of local advertising in a 5,000 household market is virtually nonexistent. There is one plumber, one electrician and maybe one painter and everybody who lives in that town knows them. Nobody will likely be willing to pay anything for Angie's service simply because there is no need for its service. Obviously no revenue will be coming from the service provider side because the one plumber in town will not be stupid enough to pay to advertise himself when everyone knows him already.
However, under the current advertising expense allocation, despite having no hope of generating any meaningful revenue from the one million 5,000 household markets, Angie's List will be able to allocate almost 20% of its marketing expense to Cohort Angie. Knowing investors do not care about the money losing markets and are only paying attention to the most mature and profitable cohorts, Angie's management would have been able to significantly understate the marketing expense in those big metropolitan markets hence boosting profitability significantly.
Instead of breaking into a large number of small markets, Angie's can decide to enter a huge market such as Shanghai. With well over 13 million in population, Shanghai probably has over 4 million household. Angie only needs to declare entrance to that market and it never needs to actually do any on the ground work to develop the market. Instantaneously, Angie's will be able to allocate a significant amount of advertising expense to Shanghai and thereby boosting profitability in its mature cohorts.
The above referenced examples are obviously extreme but it demonstrates the inherent flaws in the accounting method employed by Angie's List. The mutual fund investors who followed Angie's methodology should think deep and hard on its validity. The auditor of Angie's should request Angie's management to justify its usage. Why is Angie's List not following the most basic matching principle?
With inherent flaws in the current advertising expense allocation, I performed a basic analysis to adjust its advertising expense allocation. Instead of allocating marketing expense basing on the flawed demographic approach, marketing expenses are allocated based on actual revenue from each cohort and the results are astonishing. Based on company data from the last three quarters, if Angie's management were to use a revenue-based approach to allocate its marketing expense, the profitability in its most mature pre-2003 cohort is only half of what it is under the current allocation method.
(click to enlarge)The basic model used to derive the above presented table is attached here.
In other words, if Angie's List had properly allocated its advertising revenue using basic accounting 101, the most mature cohort's profitability would be half of what is currently being presented to investors. That is unfortunately for the long investors who blindly followed the flawed assumption put forth by Angie's management.
While the article is not intended to address the debate on Angie's business model, it is difficult not to draw parallel between Angie's misrepresentation to consumers and its misrepresentation to investors. It is also curious whether the recent CFO resignation has anything to do with the issues of the current cohort analysis put forth by Angie's. Angie's representation in its cohort analysis defies the most basic accounting principle that should be known to students in their first managerial accounting course. Such accounting method warrants additional scrutiny from auditors and regulators alike. Most importantly, for Angie's investors, think twice about the premise of your model. If an assumption does not make sense, change it to something that does.