Some Spotify Charts

Sometimes it’s helpful to turn data into another medium. It is easy to get lost in the numbers, so I often find it helpful to view the information from a distance. Charts are great for that. I am still working on my Spotify analysis, but here are some preliminary findings. I will show the chart first, followed by my thoughts.

• As I have previously calculated, Spotify’s monthly growth is around 3.738%. We don’t have enough data to tell if this is a rapid rate, since no data is available from competitors. That’s why we can only compare Spotify with itself, at least for now. Paying subscribers are growing at nearly the same rate as active subscribers, which use the service for free. I would have predicted that free users would be growing much faster than paying subscribers, but the data seems to show this isn’t the case. Currently, roughly 25% of users pay for Spotify. This seems a little too good to be true, but I assume Spotify would like to convert even more users into paying subscribers. 

• Revenues and Cost of Revenues are an extremely basic measure of profitability. In fact, these two data points sum only to give us Gross Profit. That said, they are arguably the most important indicators of growth and scale. Revenue tells us growth, while Cost of Revenue implies scale. Spotify shifted from a negative Gross Profit in 2010 to a positive one in 2011 and on. In short, this means that Spotify is making more money than it costs to power the service (without taking into account R&D and SG&A). Many startups have higher Cost of Revenues than actual Revenues, implying that they are not operating at scale. Spotify is still not profitable since the company isn’t posting Net Profit’s, but it appears to be running efficiently, as far as most startups go. 

• 91% of Spotify’s revenue comes from subscriptions, while a measly 9% comes from advertising. This is a paltry amount, considering that 75% of users are free users. It makes you wonder how little Spotify pays out to artists for every play a free user makes. No wonder many musicians are hesitant to put their music on Spotify - it barely pays. 

I often like to gauge a company’s future success by a simple question. Would I invest in it? Currently, Spotify is unprofitable, but if they can spread their costs over an increasing amount of paying users, it’s very likely that Spotify can turn into a profitable business. 

Tensions in Social Media Valuations

Startups are extremely popular today, since the Internet and the post-PC era have allowed small companies to reach an audience of billions. Everybody carries a smartphone with them, which gives the potential access to your new idea to users. Assuming your startup idea is good, you can really hit it big. We have seen this happen to Facebook, Twitter, and Snapchat, who grew (and continue to grow) users at levels retail businesses would kill for. By the way, these services are free.

After reaching such a large audience, these social media companies require money to pay for the cost of doing business, which they first get from investors in angel rounds. The investors, in return, expect to be paid back more than their initial investment. Thus, a social media company (and all companies that require investment, really) must somehow start producing revenues and hopefully profits to pay their investors back. Since their service is free, online advertising has been the only way to generate these revenues. The way online advertising works is the more users see or click the ad, the higher revenues a social media company will get. Key performance indicators (KPI’s) such as monthly active users (MAU’s) are used to calculate a valuation on many such advertising companies.

As I’ve written in the past, these new valuation techniques are extremely young, and their results are to be taken with a grain of salt. My favorite view on this issue is from Aswath Damodaran, a finance professor at NYU Stern. His take says there are currently two major ways to valuate social media companies. The first is from the traditional investor, who looks at revenues, profits, and investments. The newer valuation technique looks at user growth and other user-focused metrics like MAU’s. He argues that as a company matures, it must transition from telling stories to showing meaningful results. These results are mainly focused on the fundamentals of business, like revenues and profits.

As a social media company grows, the way it is valuated shifts from the younger model, to the older, traditional model. Most of the successful social media companies are still too young to fully appreciate this valuation shift. In fact, most investors will probably shift their techniques unknowingly. They will notice how social media companies are maturing, and begin applying the traditional models. Stories will no longer be enough to take investor’s money, and results will have to be shown.

How Fast Has Spotify Been Growing?

If you've been reading this blog lately, you will notice my interest with Spotify. It's a relatively young company, and not much data is available online about it. For the past week, I have been trying to figure out the growth rate of paying Spotify subscribers, which I have finally calculated. It has been difficult mainly because Spotify doesn't announce growth numbers every month. Fortunately, I found data online to have enough numbers to work with. Below is a table I have created listing the date and and paying subscribers that Spotify has publicly announced. I calculated the number of days and the growth percentage between the dates, and then multiplied them by 30 (number of days per month). This gives us the a very rough growth rate of paying subscribers per month, which we can use to predict future subscribers. 

As you can see, the growth rate per month (from now on just growth rate, since it's shorter to type) varies from a high of 18% in November 2011 to a low of 4% in November 2014. These numbers are deceiving, as they calculate the growth between two unequally distant periods, divided by the days between those periods. Some months may have had higher rates of growth due to well-placed advertisements, entering new markets, or just fortunate word of mouth. What these calculated growth rates did was give us a ballpark figure, since no other data was available online. So now I knew the approximate rate of growth, and now I needed to check if I was on the right track. I did just that. I had the data Spotify announced in March 2013, as well as the number of subscribers in May and November 2014. If my rate of growth from March 2013 was correct, I would be able to multiply the paying subscribers by the growth rate and come up with the actual data Spotify announced. 

The first growth rate I tried was the one from May 2014, or 4.69% per month. If this was the correct rate, I could calculate 10,000 paying subscribers for March 2014 and 12,500 in November 2014. Here is what the 4.69% growth rate got me.

10,400 users in March and a whopping 15,000 in November - Spotify wishes! This told me that my calculated 4.69% rate was too high, and I had to adjust it down. Next, I tried the most recent rate from November 2014, which was 4.08%. Again, here's the calculation.

9,695 users in March, and 13,350 in November. Again, I knew I was off. The rate per month was actually higher leading up to March, and much lower from March to November. My goal wasn't to calculate the growth for each month (that is impossible, since I don't have enough data to work with), but rather, it was to calculate the approximate growth rate per month between the periods of March 2013 and November 2014. Since 4.08% was too high, I knew I had to adjust it further. I kept adjusting the rate until I came up with 3.738%. That is the average growth rate that the Spotify paying subscriber count kept growing at every month. Again, it's worth reiterating that this is an approximate rate, and some months grew much faster than others. As you can see below, my calculations almost perfectly matched the data provided by Spotify for March and November 2014. For March, I got 10,030, when the actual number was 10,000 (Spotify likely rounded this number, so my calculation may be perfectly precise). And for November, my calculation was precisely right - 12,500 - just what Spotify announced. 

This rate of 3.738% is useful for many reasons. Foremost, it will allow me to predict subscriber counts in the future with a reasonable accuracy. With that, I can finally begin to put a value on Spotify as a whole. This analysis was mostly numerical. With the calculated data, I'll be able to dig into the real meat of the story, which is how Spotify can grow, strategies to do so, the operations underlying those strategies. 

How is Spotify Growing so Rapidly?

As I'm working on a deeper financial analysis of Spotify, I started pondering how Spotify plans to grow and differentiate itself from the other music streaming services. Based on their latest actions, it appears they are partnering with complementors (which are services that increase the value of Spotify), developing good cross-platform apps, and aggressively pricing and marketing their service.

Partnering with Complementors

There is very little platform lock-in with music streaming services. Anyone can sign up for Beats Music, use it for a few months, and then switch to Spotify. While the playlists you make on one service don't transfer to the one you switch to, it's not an issue for most users. People just want to stream particular artists and songs on demand, which all of the streaming services easily provide. Spotify is well aware of this issue, so its been smart to partner with complementors like Facebook and Uber (in addition to many more). For example, the partnership with Facebook allows you to log into your Spotify account and easily find what all of your Facebook friends are listening to. With this feature, you're able to find curated music choices from the people who matter the most in your life.

In addition to Facebook, Spotify also partnered with Uber last week, further increasing their supply of complementors. This latest partnership allows you to play all of your songs while in an Uber car. The goal here is to further increase the value-add of Spotify, as compared to their competitors.

Spotify also makes available a third-party API that allows application developers to tap into the Spotify music collection. Apps like Djay use this API to add extra features on top of the already existing Spotify service. The effect is to further lock-in users, by providing them with more value. This, of course, comes at a lower cost to Spotify, since they don't have to develop these third-party applications - they only have to build the API.

Cross-Platform Apps

This strategy needs very little explanation. Spotify wants to reach as many users as possible, so it builds applications on as many platforms as it can. It has apps on all of the major platforms (iOS, Android, Windows Phone, Mac OS X, Windows), including the web. If somebody wants to try it, Spotify made sure its service will be available anywhere.

Aggressive Pricing and Marketing

The price for music streaming services is an established $9.99 per month on all the major services. Spotify also offers family and student plans, however. The family plan is $10/month for the first family member, and is discounted to $5/month for additional members. This isn't a novel feature, but not all of the music services offer a family discount. Again, Spotify wants to appeal to as many users as possible.

There is also a student plan that goes for 50% off, or $4.99/month. The aim here is presumably to indoctrinate students, who will eventually graduate and switch to the full price plan. Students are also much more likely to download their music illegally, and having them pay discounted rates is much better than having them pay nothing. Lastly, Spotify must know how vital word of mouth is for younger audiences, which essentially provides free marketing.

It's no wonder why Spotify has been growing faster than their competition - they've been engaging in beneficial partnerships, providing access to all the major platforms, and pricing themselves aggressively. This doesn't mean that they will succeed in the long term, though, it just means they're currently doing well. Perhaps Taylor Swift was foolish to pull her music off Spotify after all?

New Valuation Techniques

I must admit, in the past I was highly skeptical of staggering acquisitions like WhatsApp, which sold for $19 billion. I did not believe that a company such as WhatsApp, with almost no revenues and an unclear business model, would be able to monetize and provide that amount of value in a million years. To be perfectly honest, I still am a bit skeptical, but I am now more willing to accept these insanely pricey acquisitions. My thinking on this topic has changed fairly recently, as I've been trying to think outside of my traditional business education, and understand this new wave of valuations. 

I'm not the only one who is trying to understand valuations in high growth industries. Here is what Jeff Liu and Jim Reinhart write in a report for EY: 

Clearly, traditional valuation metrics based on multiples of current earnings or revenues don’t accurately reflect the future expectations that the valuations imply in recent large transformative transactions. Consider the $19 billion price tag placed on WhatsApp, which reportedly generated only about $20 million in 2013 revenue — a multiple of 950x revenue! In fact, applying traditional valuation approaches may cause companies to overlook acquisitions that can transform their business by propelling growth in new markets or industries.

This whole time, I was looking at these acquisitions through my traditional accounting and finance education, which preaches metrics like book value, net income, and simple year over year growth. While these are still solid metrics to gauge most businesses by, they begin to break down for unprofitable companies that are experiencing rapid growth, which defines most social media startups. The truth is, there is no fundamental metric that can precisely predict the future growth, profitability, and success of companies in this industry. Social media, startups, the other companies that we associate with Silicon Valley all fall into an extremely young industry that has not been studied in enough detail yet. Everything in this business is an estimate, at best. And while new metrics to value these companies have emerged, they are not battle-tested enough to spend billions of dollars based on their results. I bet a large part of the WhatsApp purchase was Zuckerberg's gut.