I just bumped into this video showcasing the creation and rise of video sharing site dailymotion. In it, the company lets the market know that they will try to reach a 20 million euros revenue in 2008.
Double checking through their numbers, you find that they were doing about 667m video views / month in november 2007. Looking at Comscore numbers recently, you see that they increased their traffic by over 2 in 1 year.
So projecting a growth of 667 x 2 for end of 2008, this gives you an average of 1000 m video views/ month (33M / day).
Hence, with a revenue of 20million euros / year only on advertising, this means that 20/ (1000*12) *1000 = 1,67 euros is the target CPM for the company. If they stay flat, and don't grow, then the CPM shall be 2,5 euros.
These numbers are absolutely doable. Keep in mind that they imply a 100% of inventory sold, something you rarely see on sites, meaning that they usually sell at 50% on a yearly average (seasonal effect on months, night/day effect). So multiply the above numbers by 2.
=> minimum CPM is then 1,67 x 2 = 3,33 euros
The questions raised hence are :
1) what percentage of the inventory can be sold with advertising, meaning where will advertisers accept to go to ?
. homepage with editorial content ?
. exported video players
. branded channels
. internal pages ?
will illegal content be sponsored with advertising ? will the porn content have ads (probably around 15% ?) will private content be sponsored with ads ? your uncle Joe's wedding ? your holiday videos ?
Let's assume here that only 50% is monetizable then.
=> minimum CPM is then 3,33 x 2 = 6,66 euros.
This number is still doable, as the audience, and the brand commands a premium. And in addition we're talking here about premium or quality content that is advertised.
2) Dailymotion has announced major deals with media companies such as MTV. They are either paying a minimum guarantee (I doubt it, VCs don't put in 25m euros to pay MG to media companies... but DM has bid for football rights in France...), or have a rev-share with them. DM has an internal sales team, hence they don't pay 30% to an outside agency anymore. But they must ast least pay 33% of rev-share.
let's assume that the monetizable content is the one that commands the highest traffic and is premium content. For simplicity, let's assume 50% of that commands rev-share of 33%.
so 50%*6,66 + 50%*6,66/(1-33%) = 8,31 euros CPM.
Still doable, but getting tough. This cannot be done on page only CPMs and involves a mix between page CPM (4-5 euros), and in-stream video ads (above 10 euros) !
3) will this enable DM to breakeven in October 2008 as announced here ?
well... I'll make many assumptions here :
- 1 Mb/s of bandwidth gives you 328,5 of transferrable GB /month
- assuming a c. 50% day/night effect, you get about 165 useful GB / month.
- let's assume that because you buy lots of bandwidth, including CDNs (akamai, limelight, vitalstream...), you can buy 1Mb/s for 15 euros / month (DM pays probably slightly less)
- this means 1 GB costs 0,09 euros.
- DM encodes videos at 425Kb/s on average (press i on any video to see the info)
- Comscore says the average length of a video is 2,7 min. Hence the average size of a video is 8,4 MB.
- the cost of delivering 1000 videos (e-CPM) is then 8,4 * 0,09 = 0,77 euros.
- If on average DM delivers 1000 million videos /month, that's a cost of 770K€ / month.
- now of course bandwidth is not enough. Add servers, routers, load balancers, storage... the cost of delivering is an easy x2. Hence a monthly cost of 1,540 m€/month, or 18,5m€/year. They could break even indeed on the last days of october :)
- but we need to add in other expenditure. With 100 people (as announced in the video), we can expect a 60K€/person on average. *1,5 for social charges in France + 10KE of computer, offices and G&A/ person. that's a yearly 100*100K€=10M€ more for 2008..., so 28,5M€ in costs, excluding any MG paid externally...
Hence, in my opinion, with only 20m€ in revenues, they shall not cover the close to 30m€ in expected costs for 2008... But of course I must be completely off chart... Costs are easy to spend, revenue is harder to make isn't it, particulartly with the high CPM described above.
Optimization areas include :
- hire less people, or much cheaper (although it's a good average)
- have shorter videos than 2,7 minutes
- do not go for rev-share and sell access to DM's audience and brand for a high fee (isn't the current price at 100KE ?)
- win every lawsuit ;)
- increase CPM to start with, because there's no clear market data, and extract as much from the market as they can
By the way the reasoning could also apply to Youtube, but it will be false as Youtube benefits from the huge economies of scope of Google (infrastructure, technical expertise in scalability, sales network).
Comments and corrections welcome !