YouTube Revenue And Profitability Analysis
The world’s largest social video site is big, growing and profitable
YouTube is one of Google’s largest and most important properties which attracts more than 1 billion unique individuals each month, or around 50 percent of all internet users worldwide.
By way of comparison, Google Search attracts about 1.2 billion unique individuals each month or about 67% of all internet users.
But while it is easy to agree on YouTube’s scale and the site’s market-leading position, it is not so easy to estimate how much revenue YouTube generates - or if it even returns a profit.
We estimate that YouTube’s revenues were about USD 5.0 billion in 2013, representing 9% of company revenues. This estimate includes revenues that Google earns from video ads and text overlay ads served on video content on YouTube. Revenues earned from premium content have been excluded because they represent a very small percentage of the total revenues for YouTube.
Looking ahead we see YouTube’s revenues are increasing because of three main factors:
- Overall growth in the number of unique viewers: for the U.S. the number of unique viewers for YouTube has increased from 150 million in January 2012 to 160 million in January 2014. International growth in unique viewers has been higher, both in absolute terms and when viewed on a percentage basis;
- Increase in the number of streaming ads served: according to comScore, for the U.S., Google served 1.12 billion streaming video ads in February 2012, but by January 2014 this had increased by 158% to 2.9 billion ads;
- Increase in the average cost of each ad: Our analysis of Google’s AdWords business (which allows customers to have their text ads overlaid on top of YouTube videos), shows that Google has been managing to steadily increase the average advertising revenue per search query over the last 5 years, which strongly indicates that YouTube advertisers are paying progressively more for their YouTube ads.
When viewed on a P&L basis, we think that YouTube is a profitable business for Google: not only does the company not have to pay content acquisition costs for the vast majority of its content, but our analysis is that the streaming costs are not as onerous as one might at first think:
We can start by gleaning some interesting insights into YouTube’s streaming costs by looking at Netflix.
For 2013, Netflix reported total costs of USD 3,083 million as “cost of revenue”, which the company defines as including ‘content acquisition, content delivery, customer service and payment processing fees.’ In its 10-K filing for 2013, Netflix also says that ‘content licensing expenses represent the vast majority of cost of revenues.’ On the basis of this, we can assume that around 15% of the USD 3,083 million might represent content delivery expenses – that is, the cost that Netflix is incurring to stream video content to its customers.
This works out as: 15% x USD 3,083 million = USD 462 million, or 10.5% of revenues for streaming costs.
Separately, statements made by Netflix CEO, Reed Hastings, in October 2013 revealed that Netflix streamed a total of 9 billion hours of video in 3Q13. If we pro-rata this we can estimate that Netflix streamed a total of 9 x 4 = 36 billion hours of video in 2013. But as Netflix is growing, the actual steaming volume for 2013 would have been somewhat less than this so we will assume that it was 27 billion hours for 2013.
If the cost of streaming 27 billion hours of video was USD 463 million, then that implies a content delivery cost of about USD 0.017 per hour.
We can then use this cost per hour figure of USD 0.017 to get an initial view of the cost of video streaming for YouTube:
We know from comScore data (see chart earlier in this section) that YouTube streamed a total of 14 billion hours of video in the U.S. in 2013. If we make a rough assumption that markets outside the U.S. added another 17 billion hours (based on Google’s revenue split between US and non-U.S. as reported in the company’s 10-K filings), then we can estimate that YouTube streamed about 31 billion hours of video in 2013 which, interestingly, is about the same as Netflix (27 billion).
If we assume for now that YouTube’s per-hour streaming costs are the same as Netflix, then this would imply that YouTube’s total streaming costs in 2013 were about USD 0.017 x 31,000 million = USD 534 million.
However, we need to make two corrections to this figure.
Firstly, Netflix is streaming at an average speed of over 1.0 Mbit/s whereas YouTube’s average streaming speed is closer to 200 Kbit/s. Streaming costs do not scale pro-rata with streaming speed, but there is definitely a positive correlation. Based on our experience with the rate cards that CDN networks provide, we think that it is right to reduce our estimate of YouTube’s streaming costs from USD 534 million to USD 400 million, or by 25%.
But while YouTube has a cost advantage over Netflix because its streaming speed is lower, Netflix has an advantage because it can use edge caching to get the content closer to the subscriber, thereby reducing the average network path length between the streaming server and the user.
The reason why Netflix has an advantage here is that the company’s entire content catalogue includes just 13,000 items – about 10,000 movies and 3,000 TV shows. This represents a total volume of data of about 30 TB, which by today’s standards can fit onto a fairly modest storage server.
In contrast, YouTube’s growing user base is uploading about 300 million hours of video a year (pro rata from the 100 hours a minute quoted by Google here). In comparison, Netflix’s entire catalogue represents just 21,000 hours – or just 0.006% of what YouTube users upload in one year.
This means that our estimate of YouTube’s streaming costs should be increased, we estimate by about 50%, from USD 400 million to USD 600 million.
So, in summary, our analysis is that YouTube’s revenues were about USD 5,000 million in 2013 while content delivery costs would have been about USD 600 million, or 12% of revenues. In comparison, also for 2013, Netflix earned revenues of USD 4,375 million with USD 463 million, with 10.5% being paid in content delivery costs.
The two other aspects to consider are content acquisition costs and marketing expenditure.
As far as content acquisition costs are concerned, YouTube does not pay content licensing fees in the same way that Netflix does: Netflix pays about 40% of annual revenues to license content from movie studios and TV producers.
Putting to one side the relatively small amount of content that is included within Google Play (which is a similar service to Netflix) then the vast majority of YouTube’s content is generated by users. If these users choose to associate their YouTube account with their AdSense account then they can earn 55% of any advertising revenue that Google earns when pre-roll or text overlay ads are shown with their videos.
Because not all YouTube content partners choose to become an AdSense partner, the total percentage of YouTube revenues that is paid out to content providers will be less than 55% - we think that it is likely to be in the same order as what Netflix pays for its content, or about 40% of revenues.
Marketing expenditure has been costing Netflix about 13% of revenues in marketing costs since 2010. Interestingly, the equivalent figure for Google (whole company) is a little lower: 11%.
In summary, when viewed on a like-for-like basis:
- At an estimated 12% of revenues, Google is paying a little more than Netflix in streaming costs;
- At about 40%, both Google and Netflix are likely to be paying about the same on content acquisition: Netflix pays a fixed licensing fee up front for each content item, whereas YouTube pays a percentage of the advertising earned from a content item to the content provider, if that provider has elected for their video to carry ads;
- Netflix pays about 13% of revenues on advertising, compared with 11% for Google (average for the whole company).
Therefore, our analysis is that Google’s YouTube business is likely to be at least as profitable as Netflix.