Slashdot Mirror


Users Abandon Ship If Online Video Quality Is Not Up To Snuff, Says Study

An anonymous reader writes "The first large scientific study of how people respond to poor video quality on the Internet paints a picture of ever rising user expectations and the willingness to abandon ship if those expectations are not met (PDF). Some nuggets: 1) Some users are willing to wait for no more than 2 seconds for a video to start playing, with each additional second adding 6% to the abandonment rate. 2) Users with good broadband connectivity expect faster video load times and are even more impatient than ones on mobile devices. 3) Users who experience video freezing watch fewer minutes of the video than someone who does not experience freezing. If a video freezes for 1% of its total play time, 5% less of its total play time is watched, on average. 4) Users who experience failures when they try to play videos are less likely to return to the same website in the future. Big data was analyzed (260+ million minutes of video) and some cool new data analysis techniques used."

34 of 155 comments (clear)

  1. If a video freezes by Osgeld · · Score: 2

    "If a video freezes for 1% of its total play time, 5% less of its total play time is watched, on average."

    no shit, cause it pisses you off to sit there watching a fuzzy video of a ZX Spectrum game that the asshat somehow encoded and uploaded at 20480P and is hosted by blip

  2. so passenger ships.. by jasontheking · · Score: 5, Funny

    so passenger ships shouldn't get dodgy video playback equipment, cause people might jump overboard, even if its freezing?

    I guess I should read the article, huh..

    1. Re:so passenger ships.. by MrEricSir · · Score: 2

      Well sure, what did you think really happened on the Titanic?

      --
      There's no -1 for "I don't get it."
    2. Re:so passenger ships.. by Mitchell314 · · Score: 2

      I don't know, I forgot the ending of the movie . . .

      --
      I read TFA and all I got was this lousy cookie
    3. Re:so passenger ships.. by jd2112 · · Score: 2

      I don't know, I forgot the ending of the movie . . .

      Leo DiCraprio Dies! Probably the best performance of his career.

      --
      Any insufficiently advanced magic is indistinguishable from technology.
    4. Re:so passenger ships.. by Anonymous Coward · · Score: 5, Funny

      It hit a digital iceberg. But since the bit rate was slow and the codecs were primitive (it was still 1912 after all), the iceberg was all jagged. Due to a couple of lost packets, the bulkheads didn't get assembled completely. Oh and some asshole damaged a router that connected the forward and aft parts of the ship when he was screwing some chick in a car. Naturally, the ship split into two smaller LANs. By then the quality of the streaming Waterworld movie went to shit and people started jumping off the ship to save their sanity (though arguably, this was an impossible task). Kate Winslet got naked at some point, but since she didn't show much, everybody on the ship started searching for their porn elsewhere. This caused the data center to overheat. Wisely, the captain scuttled the ship to provide additional cooling. The End.

    5. Re:so passenger ships.. by kelemvor4 · · Score: 2

      Well sure, what did you think really happened on the Titanic?

      I stopped watching it because the video froze for a few seconds and never saw the end.

  3. BIG data? by viperidaenz · · Score: 2

    260,000,000 minutes of watched video is less than 1 day of youtube viewing (500 years per day)

    1. Re:BIG data? by ZahrGnosis · · Score: 3, Insightful

      Ah, good, I'm glad someone already mentioned this. Big did not deserve to be italicized there not only because 260 million minutes of video isn't "that much" (!) in terms of internet streaming viewers, but the statistics aren't really based on number of minutes of video analyzed... the main statistics are more about viewership and certain events (video startup, video freezing), which could be surrounded by hours of uninteresting video time that didn't really contribute to some of the metrics.

      Netflix has, what, 20+ million individual viewers per month? 10 hours a piece isn't hard to imagine. As the parent pointed out youtube is much larger than that.

      It's still very interesting analytics. it's not always the size that matters with "big" data. But let's not get carried away with the italics now people... this way madness lies.

  4. Users who experience failures when they try to pla by CamD · · Score: 2

    Users who experience failures when they try to play videos are less likely to return to the same website in the future.

    Take note, Slashdot.

  5. Romero Institute by girlintraining · · Score: 2, Insightful

    People don't keep using things that are broken, says latest scientific study from the Romero Institute. Professor Obvious, chair of the Three Kinds of Lies committee, said today that it was a shocking discovery. Many businesses have for years been selling things that are intentionally broken and assuming that people would simply keep buying them despite alternatives being available. Obvious has been nominated for an igNobel prize for his work, and says future studies may even uncover the precise mechanics behind why people continue to not use things that don't work.

    --
    #fuckbeta #iamslashdot #dicemustdie
    1. Re:Romero Institute by Anonymous Coward · · Score: 4, Insightful

      Professor Obvious, chair of the Three Kinds of Lies committee, said today that it was a shocking discovery.

      Could you find a new hobby besides posting here? The purpose of studies is not just to confirm knowledge or common sense suspicions, but to quantify that knowledge. There is no way in fuck that Professor Obvious knows a priori that an additional 1 second delay will cause 6% of viewers to flee. Professor Brilliant might know this, but that ain't me and it ain't you.

    2. Re:Romero Institute by Frosty+Piss · · Score: 4, Informative

      Look, no one gives a flying fuck about YouTube videos of Fluffy and Buffy if we're talking about cute poodles. Now, if we are talking about two-legged bitches, that's another issue.

      Let's be serious here, we're talking about porn.

      Seriously, it's hard to get hard with choppy video of the "old in-and-out". For the total turn-on, we require high quality video and a nice transfer rate. Seriously.

      I mean, how am I going to get my "freak" on with stilted choppy bad video? I might as well go down to the local "Adult Entertainment Shop" and buy a CD or check into a booth (with complementary kitchen towel roll (COSTCO) and a sticky floor)...

      Come on people, good video and "personal satisfaction" go, er, hand in hand...

      --
      If you want news from today, you have to come back tomorrow.
    3. Re:Romero Institute by aNonnyMouseCowered · · Score: 2

      "The purpose of studies is not just to confirm knowledge or common sense suspicions, but to quantify that knowledge. There is no way in fuck that Professor Obvious knows a priori that an additional 1 second delay will cause 6% of viewers to flee. Professor Brilliant might know this, but that ain't me and it ain't you."

      I think the OP's point is that some research is simply not worth the paperwork and grant money. I mean knowing precisely how broken a video can be before people stop watching is interesting, but theire are more interesting, and possibly more important, things out there. Studies like these divert resources from those research projects especially now during hard economic times.

    4. Re:Romero Institute by chmod+a+x+mojo · · Score: 4, Interesting

      Well I think that video streaming sites would be VERY interested in this data. Probably interested enough to at the very least partially fund the research.

      Hmmm, Thinking about it, probably any type of retailer would be interested in data like this. It's a quantifiable amount of time before loss of interest, not just "customers hate waiting".

      Another interesting tidbit from this study, it's probably a bad idea to put an ad at the very beginning of the video ( for ad supported sites ) since most ads are more than two seconds long. This may seem counter intuitive since if you show the ad BEFORE the video you shouldn't have to interrupt the actual video - like hulu does it - and you would think users would prefer getting it out of the way first so as not to be interrupted. Then again that breaks the traditional commercial model that people are used to from television and may take them out of their comfort zones.

      --
      To err is human; effective mayhem requires the root password!
    5. Re:Romero Institute by gnoshi · · Score: 2

      This study is really a revelation for me, the findings are highly non-obvious. I had thought that people would wait indefinitely for the video to appear, based on the "sunken costs" theory, i.e. I've already invested mm minutes and ss seconds waiting for this to appear and it might appear at any moment.

      This actually captures well why people should hesitate before deriding studies which have seemingly obvious outcomes. This study may be on the margins of that - although it is the quantification that is actually interesting about it - but sometimes studies find counterintuitive results. Even better, if a study produces a what may be a counterintuitive result then hindsight bias means people will tend to revise memories so they believe that was the expected outcome all along.

  6. less about quality, and more about functioning by Trepidity · · Score: 5, Insightful

    The metrics mentioned aren't really about video quality, which I tend to think of as things like the resolution, encoding artifacts, sound/video sync, etc. These are more about the video player functioning correctly, at any quality of video: that it starts playing the video soon after the user hits "play", and it doesn't drop out during the middle of playing. That's a kind of video quality, sure, but it's closer to "I stopped watching b/c the damn player didn't work" vs. "I stopped watching b/c the video's quality was too low".

    1. Re:less about quality, and more about functioning by ArhcAngel · · Score: 2

      This isn't about quality of the video at all. It's about QOS for video.

      This report is what Comcast uses to determine just how much to throttle Netflix to get the most people to come running back to cable but not run afoul of the FCC.

      --
      "A person is smart. People are dumb, panicky dangerous animals and you know it." - K
  7. Five... by roc97007 · · Score: 5, Insightful

    5) Users bail when the video loads and it's a commercial that can not be skipped.

    Because unwanted, unskippable commercials are exactly like a pause before the video starts equal to the number of seconds the commercial plays. (See (1).)

    --
    Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
    1. Re:Five... by Splab · · Score: 2

      What really pisses me off is when it starts playing automatically which a lot of US news sites do.

      Oh and commercials that are longer than the video content. Or commercials for pure US products...

  8. Top 10 Online Video Complaints... by dtjohnson · · Score: 5, Informative

    10. "You don't have Flash 10.7 installed and need to upgrade to Flash 10.7" when you're running Flash 11.x
    9. Embedded ads
    8. 'special' video players (I'm looking at you ABC)
    7. Video freeze during play due to lack of server response
    6. Sound but no video
    5. Video but no sound
    4. Incompatible video formats
    3. Video resolution inappropriate to the method of delivery...either way too high or way too low
    2. Websites that insist on posting useless bandwidth-hogging 'talking head' videos rather than posting a simple photo and a text summary.
    1. Digital Rights Management and all its limitations

    1. Re:Top 10 Online Video Complaints... by xetovss · · Score: 5, Interesting

      You forgot to put this on your list: Videos skip and pause when fully buffered . I am not sure what is the actual cause for it is but something causes it, even on a 2.8GHz Core2Duo w/ 8GB of memory with Win 7 64bit I get that a lot especially with youtube sometimes with others.

    2. Re:Top 10 Online Video Complaints... by Gaygirlie · · Score: 2

      2. Websites that insist on posting useless bandwidth-hogging 'talking head' videos rather than posting a simple photo and a text summary.

      This one is one of my pet peeves: I can't even count the number of times I've wanted to know more about something that could have been explained much more clearly AND in only a few passages of text but the author(s) insisted instead on making a video with a talking head that adds absolutely not a single thing to the topic. What's the point? Video should be actually used for something, but when it's just a head saying the same thing that could've been said in written text it's waste of everyone's time.

    3. Re:Top 10 Online Video Complaints... by adolf · · Score: 3, Informative

      I had this issue with recent FF builds under Windows. With one simple change, it works as fine now as it did several years ago:

      Tools > Options > Advanced > Use hardware acceleration. Uncheck* this. Press OK. Restart FF.

      Done.

      *: Yes, really.

  9. Re:Brilliant business model. by sycodon · · Score: 4, Insightful

    Ever notice that the Advertisements load faster and are of better quality (DPI) many times than the video?

    --
    When Fascism comes to America, it will call itself Anti-Fascism, and tell you to give up your guns.
  10. Pfft video by megrims · · Score: 5, Insightful

    Personally, I bail when the content is a video. Give me back my plain text internet, please.

    Videos are such a waste of time.

  11. Re:Brilliant business model. by Anonymous Coward · · Score: 3, Informative

    In fact, I have! I live in China and tried to use Hulu through a VPN. The actual shows SUCKED(lag, stutter, failure of the player to transition from ads back to the regular show) but the ads never skipped a beat.

    Been using pirate bay and have never looked back.

  12. Reading the stats wrong by __aaltlg1547 · · Score: 2

    They're not accounting for the 10 times a day I bail on a page because I DON'T WANT TO WATCH THEIR STUPID VIDEO AT ALL.

  13. Re:Brilliant business model. by runeghost · · Score: 4, Insightful

    Not exactly. What I notice about ads is that they often try to load in a higher quality than the video I'm watching, then stutter and choke on the crappy bandwidth that is the best I can get where I live. Or they try to do something fancy and interactive, and hang or crash my browser. And then I wonder again why I'm not just downloading my content from the pirate bay...

  14. Re:This again by bmo · · Score: 2

    You write this as if Google Analytics is the only ad service out there.

    Google Analytics is pretty responsive. Very responsive if you compare to a lot of other ad services.

    Another thing that has bogged down browsers over the years is the sheer amount of JS that dwarfs the amount of HTML on a web page, leading to an arms race to see who wins - the web page devs or the browser JS engine devs. The engine devs lose most of the time, and when the rare occasion happens that there is a breakthrough in JS speed, web devs tack on even more crap, because they can. Add to this the ridiculous amount of JS that ad services and trackers throw in, and you have a fucking nightmare of inefficiency. Adblock is the only real way to combat this for now.

    But this has nothing to do with video playback once the stream is started. Once the ads are loaded or blocked locally, the video itself should play and the server should keep the buffer full for the player, which often times fails. QOS for video is atrocious from the viewer's POV. While it's amazing that Youtube is able to stream as many videos as it does without completely imploding, this does not matter to the viewer. What matters to the viewer is that the video starts playing, and then a minute or so into it, the buffer goes away, because the video server forgot all about the stream and "will get back to it in a bit." Then the viewer bails unless it's really compelling and he hits pause and waits for the buffer to fill again, or the viewer(a very small percentage of the total) uses a downloader to save the file locally and later viewing, which I do sometimes (I do this for videos on wimp.com).

    From the viewer's POV, video on the Internet sucks for the most part. It's not good enough for non-casual watching and is unlikely to be so for quite some time unless infrastructure becomes better.

    --
    BMO

  15. Re:Brilliant business model. by SeaFox · · Score: 2

    No, in fact I see the opposite on Hulu quite often. The advertisements are of such poor quality I sometimes wonder how Hulu tricked companies into paying for the placement.

  16. Re:Just press pause by the+eric+conspiracy · · Score: 2

    The problem is more fundamental.

    The internet is fundamentally a best effort statistically switched packet network. No delivery guarantees. No particular order of delivery.

    Video uses temporal compression and requires timely delivery for a stable reconstruction of the video. Drop a master frame and all hell breaks loose. At a low level this is incompatible with the design of the internet.

    Throw in the fact that people are conditioned to a highly reliable delivery system (cable TV) with dedicated bandwidth and you are simply going to find that they are not going to tolerate the performance profile you are going to get trying to cram a high bandwidth stream down a statistically switched pipe.

  17. Re:Brilliant business model. by smellotron · · Score: 3, Informative
    Ever notice that the Advertisements load faster and are of better quality (DPI) many times than the video?

    No

  18. We want TV but not TV. by jellomizer · · Score: 3, Interesting

    So in short, We want TV Quality Video.

    Not so much news. If the video is choppy or looks bad, we tend to not want to watch it.

    There is the people who called Color TV a fad. However its success was in the fact that the Color TV didn't come with a bunch of disadvantages, It was better to have color vs. Black and White. Now with Internet Video. There are advantages to it. However Lag and Quality are major disadvantages. And will not catch on unless both are resolved.

    In many cases in both Lag and Quality have improved with advancements in network speed combined with better quality data compression, However still the load times means we need to invest into watching something vs. the old flipping through channels, to see what is on and if it catches you attention.

    --
    If something is so important that you feel the need to post it on the internet... It probably isn't that important.