|
Pioneered by WarpPrism, serious contenders for alternative stream browsing client/aggregate have popped out from the likes of www.Twitch.TV and www.WellPlayed.tv to streamline and add value to the viewer experience. Although almost every option serves Justin.tv streams, each client have built their own unique sets of features, but two that all share (or will have) are fast stream switching and Picture-in-Picture (PiP).
One unintended (or just largely overlooked) consequence of these two features are its financial repercussion with the casters whom provide the entertainment. Fast-switching provides a "channel-flipping" like experience for the viewer, and to enable it one necessary sacrifice to make this possible is to disable pre-roll ads, an important component of caster revenue. PiP, on the other hand, enable viewers to practically watch multiple streams in a simple browsing interface, created a sudden, effective increase in viewer base from which draw ad-impressions from. These two variables moves revenue streams in opposing directions, whether one willl overpower the another, and the magnitude of these change, can have significant repercussion to this significant source of income in our nascent industry.
Terms - Impression: one instance where a viewer watched an ad from start to finish
- Pre-roll: A term for ads that played when first tuned into a stream. It can be considered a "free" impression and has significant correlation with unique visitors (for the purpose of this study, they're consider equivalent).
- Standard ad: Streamers can run an ad on Justin.tv every x-minutes they streamed. One important feature that Justin.tv provides to game-casters is the ability to "stock-up" on ads and choose when to run them. An effective streamer can run appromixately 3-4 ads a hour.
- Ad revenue: roughly $2~5 per thousand impression according to various sources.
Pseudo equations - Impression = Pre-roll ads + Standard ads
- Revenue = Impression * $$/thousand
Variables: Let- M = unique viewer multipler (define: unique viewer / average viewers)
- V = average viewer
- WP = a number between 0 to 100% that corresponds to the percentage of viewers that utilize alternative video clients
- PiP = a number between 0 to 100% that corresponds to the percentage of alternative client users that uses PiP
- Ad = average ads played per hour
- Hour = average hours per session
Equilibrium points: The pseudo equations collapse to:
- Impressions = M * V * (1- WP) + (1 + WP) * V * PiP * Ad * Hour
- Revenue lost due to pre-roll block = M * V * WP
- Revenue gained due to PiP effective viewership increase = WP * V * PiP * Ad * Hour
System is equilibrium if and only if revenue lost = revenue gain:
- M
V WP = WP V PiP Ad Hour
- M = PiP Ad Hour
System is benefitial to casters if revenue lost < revenue gain
System is detrimental to casters if revenue lost > revenue gain
Interesting enough, because we have removing a substantial number of variables, whether the growth of alternative clients have an affect on ad revenue is fairly simple and dependent only a single variable - the viewer multipler M. PiP, Ad, and Hour are all fairly basic variables that can be reliably approximated.
For example, in a typical 3-hour stream, a M value higher than 10 guarantees that growth of alternative stream client will lead to a substantial decrease in revenue, whereas a M value lower than 3 guarantees substantial revenue growth.
Conclusion This system of analysis lead to a surprisingly simple conclusion that depends on a single measurable variable M. Current trend indicate that alternative stream clients are popular and their usage growing, thus understanding where does M stands have a paramount importance to how client developers, stream providers, and us viewers should welcome or discourage these clients and its features. I welcome Justin.tv and its streamers to do studies to determine the exact M value of SC2 streaming, thus we can all understand the proper stance forward.
Further study How to determine M, where M is the unique viewer multipler defined as unique viewer / average viewers.
For streamers
Stream multiple sessions of similiar viewership level, set half without pre-roll and half with pre-roll. The difference is the total pre-roll impression, which when divided by your average viewership will beget M.
For stream provider
M is the divisor of unique viewership of a stream against the average viewership.
  
|
Well, according to the movie by the guy who's making warp prism, Justin.tv is actively working to get pre-rolls ads working on alternative clients. If they do, there's really no loss in income for justin.tv and the streamers.
|
Some additional info is that most of our players stream for a couple hours at once and have about 10-30 minutes of viewer time per viewer. This means that per viewer preroll ads play a large role.
I would also assume that once a game finishes some people switch out to another channel to see if anything is going on there, thus missing the ad completely on their original channel.
On July 11 2011 20:44 Tobberoth wrote:
Well, according to the movie by the guy who's making warp prism, Justin.tv is actively working to get pre-rolls ads working on alternative clients. If they do, there's really no loss in income for justin.tv and the streamers.
Doesn't that hurt the whole experience of fast switching?
|
I thought ads gave .02 cents per view?
|
On July 11 2011 21:31 Dusty wrote:
I thought ads gave .02 cents per view?
that's what he said.
0.02 x 1000 = $2
|
On July 11 2011 21:19 Liquid`Nazgul wrote:Some additional info is that most of our players stream for a couple hours at once and have about 10-30 minutes of viewer time per viewer. This means that per viewer preroll ads play a large role. I would also assume that once a game finishes some people switch out to another channel to see if anything is going on there, thus missing the ad completely on their original channel. Show nested quote +On July 11 2011 20:44 Tobberoth wrote:
Well, according to the movie by the guy who's making warp prism, Justin.tv is actively working to get pre-rolls ads working on alternative clients. If they do, there's really no loss in income for justin.tv and the streamers. Doesn't that hurt the whole experience of fast switching?
This is the most interesting part I found about this analysis. The variables are more or less constrained by the revenue system we have to work with, which means with some small amount of time researching the actual values (or range of values) of M, PiP, Ad, and Hour, we can easily figure out in what direction will ad revenue will be shifting.
From the sound of your experience, Nazgul, M is awfully high, which means that there's an almost certain chance that a proliferated alternate stream client is going to hurt the players' bottom line. This means that for client developers like WarpPrism, WellPlayed.tv, and Twitch.tv, there exists an inherent conflict of interest between a better viewing experience versus player revenue.
I wonder how this will be resolved?
|
I guess the ultimate goal for the alt clients is to put their own ads, which is a catastrophe. JTV (and co) either have to somehow prevent alt clients from using their content or the alt clients have to let the viewers see jtv ads.
I know I'm not using adblock on streaming sites (or TL :p) to support the players and I hope a solution will be found about this, because if one thing is hurting ESPORTS for real, it's this :/
|
Northern Ireland2557 Posts
Realistically the pro's are gonna get shafted due to the overwhelming majority of viewers compared to them. Majority rules.
|
On July 11 2011 22:06 MrCon wrote:
I guess the ultimate goal for the alt clients is to put their own ads, which is a catastrophe. JTV (and co) either have to somehow prevent alt clients from using their content or the alt clients have to let the viewers see jtv ads.
I know I'm not using adblock on streaming sites (or TL :p) to support the players and I hope a solution will be found about this, because if one thing is hurting ESPORTS for real, it's this :/
i think if the other clients started shafting jtv for money then jtv would stop allowing them to restream the content
|
On July 11 2011 21:36 me_viet wrote:that's what he said. 0.02 x 1000 = $2
0.02 x 1000 is 20.
Mathematically you get 0.2 cent or $0.002 per viewer.
|
I know I have noticed a drop in our ad revenue in line with how much our traffic from WarpPrisim has grown.
<3 WP, but this pre-roll ad issue needs to be fixed T_T
On July 11 2011 20:44 Tobberoth wrote:
Well, according to the movie by the guy who's making warp prism, Justin.tv is actively working to get pre-rolls ads working on alternative clients. If they do, there's really no loss in income for justin.tv and the streamers.
They would start after 15 seconds or something. Might be a good middleground, but only time will tell.
|
JTV should really follow livestreams method of showing pre-rolls - run it after 20 - 25 seconds so you can figure out what is being streamed and whether you want to keep watching. Currently channel surfing on JTV without a 3rd party stream client is pretty much impossible with all the pre-rolls enabled on 95% of channels and it really encourages use of adblock.
|
Originally, I started working on this model to prove that the clients are a net positive, so the conclusion was a (pleasant) surprise just from a theoretical prospective.
Personally, I don't want to give alternative client developers too much shit for what they do, for they too are creating value in this nascent industry. More importantly, now that WP has opened the pandora's box, viewers are not gonna go back to the way it was. It's a dynamic shift in viewership habits and rather than complain and hate, we should be considering how to take advantage (or at least mitigate) this development.
Hopefully that rumour that JTV and WP is working together to keep pre-roll is true. If so, I am more than happy that my model can be completely discarded.
EDIT: http://www.reddit.com/r/starcraft/comments/iq7e4/warp_prism_now_with_ads_a_good_thing_it_supports/c25t94h
|
|
|
|
|
|
TeamLiquid StarLeague 5 in 
EPT
