Charbax.com

October 22, 2009

Google implements my idea for Google Reader Recommendations

Filed under: Ideas — Charbax @ 9:15 pm

I have been posting this as a feature request on Google feedback groups since May 2007, today Google is launching real feed item recommendations, which they call Magic.

My request from May 20th 2007: Backup up somewhere cause I cannot find it in the archives of the official Google Group

Subject: Personalized News Items Suggestions Actions…
From: Charbax ()
Date: May 20, 2007 7:14:39 pm
List: com.googlegroups.google-reader-past
We can star items, now all Google Reader needs to provide is a
“Recommendations” button, which would display a list of recommended
news items, based on comparing this user with the activity of all the
other Google Reader users.

There should be some settings for different kinds of Recommended news
items:

x Generate Recommendations Based on read news items

x Generate Recommendations Based on starre news items only

<------------> A slider to choose level of obscurity that news items
may be, thus only fetching the most popular other news items, or also
recommending some more obscure stories that only a few other Google
Reader users have read or have starred.

Some other algorithms and functionalities might be chosen in the
settings of this feature by the users.

And as there is a Star function now, it is important to also have a
“Discard” item button, which basically tells the Google Reader
Recommendations engine that it should please try to avoid recommending
this type of news item in the future.

A history of chronologically starred and discared news items should
also be consultable so as to maybe later make some changes in those
ratings.

Also this could be used to Suggest similar news item that is better
suited than the one that is being read. Basically, Google knows all
the related news items that talk about the same things, so using the
stats of starring, it could display a “Hey, try reading this version
of the news item on this other blog, it should be a better one and one
Google Reader thinks is of your personal taste”.

Google should contact me if they are going to steal my idea, or I have
lots of other similar visions, I’m sure they are in the works and
studied, but maybe they are not..

In December 2007: http://groups.google.com/group/google-reader-feedback/browse_thread/thread/655e44352560e9ab/

And a couple of months ago: http://www.getsfn.com/google_reader/topics/feed_item_recommendations_based_on_likes

October 9, 2009

My new speculation on Youtube’s potential

Filed under: Ideas, Video-On-Demand — Charbax @ 12:27 pm

I like to speculate on Youtube’s potential for profits and revenues on my blog in those previous posts: http://charbax.com/2009/04/09/what-google-pays-for-youtube/ , http://charbax.com/2008/08/14/current-global-youtube-bandwidth-might-be-126-petabytes-per-month/ , http://charbax.com/2008/08/07/google-should-activate-overlay-youtube-ads-now/ , http://charbax.com/2008/06/28/youtube-needs-to-change/ , http://charbax.com/2008/04/30/when-google-starts-to-revolutionize-youtube-using-overlay-advertising/ and http://charbax.com/2008/01/09/i-just-interviewed-the-youtube-founders/

Today Youtube announces officially that they are serving more than 1 Billion video views per day. So here are my latest calculations posted to: http://www.techcrunch.com/2009/10/09/youtubes-new-logo-shouts-from-the-rooftops-1-billion-views-per-day/#comment-3027990

Let’s calculate:

1 Billion views per day, if average 30 seconds per view, if average is 300kbit/s video bitrate = That’s “only” 1125 Terrabytes of video streaming bandwidth per day.

Let’s guess Google pays $0.01 per GB of bandwidth (because they own most of their bandwidth infrastructure caching all videos as close to viewers as possible). That would mean that Google would be paying only $11250 in bandwidth per day.

If average Youtube video view is 1 minute, that would still only cost $22500 per day at $0.01 per GB.

If Google pays as much as $0.05 per GB, and average video length is 1min and average bitrate is 300kbit/s, that’s still “only” $112500 per day to host the worlds biggest video site reaching hundreds of millions of viewers all over the world.

Now let’s consider Google may be “only” monetizing 20% of those views at a rate of “only” $10 per 1000 views with overlay advertising in partner/copyrighted/claimed videos:

If 20% of Youtube videos are monetized at $10CPM, that’s $2 Million in revenues for Google per day, for 200 million overlay and other ad impressions per day.

Overall, according to my calculations, Google may be spending between $8.2 Million and $41 Million per year on bandwidth to deliver all those video views.

And according to my calculation and speculation, if 20% of video views are monetized by overlay and other ads, Google could be making upwards $730 Million in revenues per year.

$730 Million -  $41 Million = $689 Million in profits for Google on Youtube each year. Of which they split about half to the content providers.

Youtube could still be a very huge source of profits and revenues for Google, and my calculations are probably much lower than Google’s really potential for monetization with this.

I believe even user generated content can be monetized, more and more advertisers only care if they can sell stuff to the viewers through those overlay ads.

I also posted this comment on the official Youtube blog: http://youtube-global.blogspot.com/2009/10/y000000000utube.html?showComment=1255103825750#c9092045352321286663

3 years, why can I still not even apply to become a Youtube partner just because I am not a resident of USA, UK, Germany, France, Japan, Spain, Canada, Australia? How about the hobbyists of the 200 other countries of the world, are they not allowed to turn their hobbies into a real business?

All this time, still no real way to sell things with Google Checkout or whatever directly from within the videos. Still no automatic subtitles generation and automatically translated subtitles in all languages. Still no regular video-downloading support! No Micro-payments for monetization of really high quality streaming and downloading!

The video recommendations algorithms still suck. No really good platforms to watch Youtube on a set-top-box. No desktop uploader with resume and maximum upload bandwidth features, no way to access ones originally uploaded videos!

3 years, still no way to re-upload better quality of videos, still no easy way for regular people to claim ownership of video and audio content when other users pirate your content. Still no way to edit out the music of a video with your own alternative legal music.

3 years, still no alternative to the crappy Flash format!

OK, I appreciate that the bandwidth that Youtube provides for 1 billion daily views is absolutely amazing work in terms of bandwidth management and hosting. Though seriously, you got 1.65 Billion dollars for it, Google has the best PHD engineers in the world, you are a lot of developpers who should be able to improve things even faster at Youtube. Seriously, get to work now and release these features! You actually have the responsability to make these features work now.

Powered by WordPress