Citi Analyst Mark Mahaney is again doing some prognosticating about the profitability of YouTube according to a new report he's due to put out today. Back in January when he did this I posted a reply that I think YouTube would hit $2B this year when he said $1B. So now it's time for round two...ring the bell!
The major issue I have with Mr. Mahaney's predictions (nothing with him personally as we've never met) is his reliance on display ad CPMs for his predictions. Last time, he compared YouTube CPM to MySpace, that social network which is slowly sliding into the quagmire of oblivion. This time it seems like he's still using display ad CPMs. So I figured I would pull in some big guns this time round to give my analysis more weight.
Here's Mr. Mahaney's prediction:
So he's predicting $1.32B now for 2011 and $1.695B for 2012 based on slowly rising CPM and year-on-year growth. This is an increase from his last prediction which said 2011 would be $1.134B based on a CPM similar to MySpace which I thought was a silly comparison. He also adjusted his revenue share 2% lower and lowered his growth predictions for the company.
Here's his previous prediction from late January:
Versus My Predictions
Previously I cited some other research to support my closer to $2B prediction for YouTube revenue for 2011. Here's a quick look
- IAB/Bain Research Digital Pricing Research 2007 - $43 CPM average
- Tubemogul Average Video ad CPM in 2008 was $12.39
- Razorfish Media June 2009 - Average CPM for video ads $15-20
- Advertising Age - Yahoo! set for $3.04B ad revenue for 2010, AOL $890M, Facebook just below that and MySpace only $360M
This time, I did one better. I went to the source and asked, as I pinged SpotXchange about this and asked for average eCPM for pre-roll video ads on the exchange and they were more than happy to give up the goods.
SpotXchange's run of network pre-roll inventory is being auctioned at a range of $5 - $8 CPM, and behaviorally-targeted and performance-based campaigns using our Otto™ technology are seeing a range of $8 - $20 CPM.
-- Mike Shehan, SpotXchange CEO
You remember Otto right? I wrote about him some time back. That's the auto-optimizing algorithm for online video ad campaigns (and soon to be world ruler when the machines rise up to crush humanity) . That's a pretty massive increase in CPM (Otto FTW!). I also asked if they cared to comment on whether or not they thought YouTube was profitable, they chose not to. A shame, but not everyone is as bold as I am I suppose.
In my previous calculations I looked at some slightly higher CPMs:
...his CPM for YouTube is stable at $1.08 and I have to believe that it's probably higher than that considering the demand for the content. At $1.08 for 700B views that's about $756M. At $2 CPM that's $1.4B in revenue, at $2.5 CPM it's $1.75B, etc (slightly more than what Google paid for YouTube back in 2006).
The 700 billion views came straight from the YouTube blog. Now, thanks to SpotX we can apply some new formulae to the mix:
|$5 CPM||$8 CPM||$20 CPM|
|YouTube 700B views||$3.5B||$5.6B||$14B|
|YouTube 350B views (50% of content)||$1.75B||$2.8B||$7B|
* I used 65% since that's about standard in the industry from what I have personally seen though I have seen it go up to 50/50.
I like those $8 CPM numbers since they showed up in both the non-optimized and the Otto-driven numbers from SpotX. So if 50% of YouTube content is showing that type of premium in-video ad with an $8CPM it could generate a rather large amount of revenue. Really to be fair I should have taken the middle ground between the numbers which would have been $12.50, but I didn't want people thinking I was overstating things.
Some Final Adjustments
I went and found some other interesting stuff. For example,shows stats on what partners might make based on what some channels are probably pulling down and some estimated earnings rates.
I also did a little further digging about and found none other than Nalts (sup dude...) who has all manner of information about How Much Money YouTube Partners Make (which made me laugh several times) in which he states:
Google/YouTube is usually paid by an agency or media buyer a CPM (cost per thousand, say between $5 and $25 dollars per thousand views)
Now that's not all ads on YouTube, that's the premium ones and not the "garbage Adwords/Adsense ads" as Nalts puts it. No one can openly say what they make there so that means we don't really know the rev share percentage either.
We need to take into account the fact that a majority of partners are probably not showing the premium in-video ads and are showing the low Adsense stuff. I took an average of some Adsense accounts I have access to and over the last 6 months the eCPM on those sites were about $1.53. These sites are no where even close to getting the same amount of traffic or being in high demand like YouTube would be so using any CPM remotely close would be a drastic understatement to me.
So say that only 10% of YouTube content is showing the premium ads at YouTube and are pulling in the $8 CPM while 100% of the content is showing display ads and getting let's say $2-3 eCPM for that shall we? (Just because ads are shown doesn't mean they're sharing the revenue with someone).
700B video views * $2CPM = $1.4B
Now say 10% or 70B views has premium ads at $8 CPM 70B * $8CPM = $560M
Total that up and you get $1.96B in total revenue. Just about the same number that I predicted earlier in the year. So I stand by my claim that YouTube could hit $2B in revenue this year and that was not even taking into account any sort of growth from the last 12 months.
If we add in 15% year-on-year growth from Mark Mahaney's chart then 700B turns into 805B video views for the year and the calculations look like this.
805B * $2CPM = $1.61B
80.5B * $8CPM = $644M
For a grand total of $2.254B in total revenue for the 2011 calendar year for YouTube. Say they keep only 65% of that due to revenue sharing and you get $1.47B which means Google could recoup the cost of YouTube again this year (because I believe they have already done so at least once).