Streaming vs File-Based Media
Image via Wikipedia
I’ve been a huge fan of the streaming model vs the file based model for as long as I’ve been thinking about digital media. I wrote this about streaming vs files in the digital music business last summer:
File based music has to exist today because there is no good mobile
broadband internet solution. When you are at the gym, in the car, on a
run, you have to have files to listen to. You can’t get your music via
a stream when you are on the go. But that’s going to change and I
believe within five to ten years, we’ll be listening to last.fm, the
hype machine, and rhapsody in our cars and iPods and phones. And when
that day comes, owning files isn’t going to be necessary anymore. It’s
not necessary anymore in my home. I have several hundred gigabytes of
mp3s (all acquired legally by the way) in a server in my basement. We
barely ever listen to them. Streaming music is better because it’s
abundant. I don’t own all the music in the world on my server. But
almost every song ever recorded is on the Internet somewhere.
I think this is even more compelling in video where the file sizes are larger, take longer to download, and eat up more hard drive space. And clearly monetizing streaming media is a lot easier because ad insertion can be done in real time which means it can be targeted, tracked, and measured.
And it seems that the market is starting to move to streaming and away from file based media. Ars Technica reports that:
with the rise of Hulu, YouTube, Veoh, the BBC iPlayer, and many more,
it’s streaming traffic that now generates tremendous concern, even as
P2P drops off in some cases. The shift, should it become a permanent
trend, is good for everyone.
The part of that quote that really got my attention was the comment that P2P "drops off in some cases." Ars Technica quotes several sources in that post, including this one:
at PlusNet [a british ISP] P2P traffic has dropped from an average of 13.4TB a day
last year to 12.2TB a day this year, and now makes up only 25.9 percent
of total traffic
The Ars Technica post also has a lot of data about the rapid increases in streaming traffic, but I think we all can see that happening right in front of us. The real insight is that streaming takes away the need and the desire to pull files from the P2P networks.
This is a great thing for everyone. Streaming is easier for users and a mainstream activity where P2P is not. My daughter, who is 17 and totally technical and at ease on the Internet, had to ask me last weekend how to download a torrent. And streaming works much better as a business model for content owners and media companies.
We are finally getting to the point where content owners are embracing the Internet, putting their content up in streaming format, and getting the financial and promotional advantages of doing that. And in the process putting a dent in the file based media business. It’s about time.
Makes sense. Think this point gets overlooked in all the cloud computing talk, the browser has the potential to become a media platform as well.
Totally agree.Streaming fits our needs: on demand, portability, minimal option limitations & instant gratificationWe’ve just been waiting on the technology and infrastructure to catch up.
Why don’t you just stream your own files? The problem I have with streaming is that I enjoy a lot of “rare” material from my favorite artists that is not available for streaming. Also, some stuff is available from some services and not others, so you need multiple sources. Not to mention that I compose and record some of my own material, so I like to have that handy.If streaming ends up being the way to go in the future, there will need to be some way to aggregate streams and/or combine it with file based approaches. Maybe a way to join in different libraries. Some kind of hybrid model certainly seems logical. The popularity of Pandora on the iPhone speaks to that.
I do stream my own files. I post a song a day to tumblr that can be listened to in playlist format by clicking on the black banner at the bottom of this blogI also have uploaded my entire library to the web at anywhere.fm
I like that concept. At some point the mobile device can provide a cache for music subscription (analagous to Google Gears applied to Google Reader).
“streaming works much better as a business model for content owners and media companies”Yes…it does. But the business of online video(enabling/serving) is still in question. You pointed it out as well: http://www.avc.com/a_vc/200…I read this interview a couple of weeks ago on TC: http://www.techcrunch.com/2… ?Some points that Jeremy Allaire made:#1 “… the concept of a long tail video market simply hasn’t become a viable business.”#2 “..a “democratization of video” that hasn’t happened, largely because of proprietary formats and licensing issues that have plagued the industry.”#3 “…the idea of for-pay online content has been largely a failure …the market for this hasn’t materialized. …and we probably won’t see a wider adoption until the consumer electronics industry breaks down the aforementioned barriers.I feel like the long tail market hasn’t materialized because there isn’t enough going on to enable the mid-tail of the video content market. Google Video for Enterprises somewhat validated that yesterday…there is a big mid-section of the market that is producing professional content and Brightcove, et al are too expensive, and YouTube et al are too commoditized. So the long-tail is still an extremely viable business(eg/ nextnewnetworks), but it’s just that the tail doesn’t include the folks we thought it’d include.”Democratization” hasn’t happened- not because of licensing and formats, but because the democracy just doesn’t have flexible, cost-effective solutions to publish brand-able/monetizable content.”Pay for content” – I’m not sure about this. I believe paid content can be a viable model(look at the porn industry)- it’s just a matter of making sure that the economics of it work. You can’t be charging your viewers a $19.99/month fee to access your content(and not know how many users you’d have in a month), and spend a huge fixed monthly cost for application platform + content delivery to support that.Just some fleeting thoughts..what say you?
I see the entire notion of “ownership” changing as internet access becomes more ubiquitous. Consumers will ideally want to own an asset and have it available across any connected device; therefore “streaming” would be preferred as the notion of having to download and manage files would become unnecessary. Consumers will be clamoring for this more as behaviors online change to adapt a free-on-demand model, but outstanding rights/legal issues will hamper big media’s ability to feed this model.
Streaming matters because it is on demand, supplants broadcast (for non-Live), and puts viewers in control.In my opinion, the laggards in the system are the advertisers and content owners who are still trying to figure out how to deal with the loss of control that strict schedules and interruptive mass advertising provided them. Viewers want control … that is why we have DVRs, skip commercials, and stream. Forced ads in direct streaming are even more annoying than on broadcast TV.The solution must include greater viewer control of the experience, even the advertising experience, such as semi-passive, opt-in, and other methods.Short-tail, Mid-Tail, Long-Tail … streaming video must find better ways to monetize the viewer-driven experience than applying broadcast techniques to streaming.Lastly … where is the value-add beyond advertising for the viewer that streaming and an interactive channel to the viewer can provide? Will the solutions be driven top down from content providers … or bottom up from the viewers themselves?Fred, I would like to follow up with you on this offline if possible.
Send me an email. The contact link on the upper right does just that
Streaming is a great concept and offers tremendous long-term value. But only if bandwidth really is ubiquitous and cheap. If moves such as Comcast’s bandwidth caps become the norm – and continued metered usage of mobile remains so – then file-based media is more cost-effective in the long-run. Better to pay only the first time we download content than every time we use it.
I have to believe that some isps will offer unmetered service and they will take share as a result. You can’t put the genie back in the bottle
I wonder when/if the rise in streaming will collide with the Internet usage caps that Comcast just announced (and others like TW are considering.) I am active Rhapsody, Hulu streamer and must consume a fair amount of bandwidth each month legally.
Agree. Bandwidth is a major infrastructure problem for the internet moving forward. If you spend enough time in a corporate network, you can see already how online video usage hampers network speed. If streaming goes mainstream (ie moving towards the extent that tv is), the cost or the speed would become detrimental. As it stands, we’re looking at more than 10 years I would say. In the end, whether it is file or stream, true “on demand” is what the user wants.
True. But: being able to download is good for the kinds of mashup and manipulability we want/need today. (Not to mention that not every device will always be able to get a network connection. Bandwidth, infrastructure, various travel issues, etc. …)
The problem with streaming is that the lack of ownership may actually increase the digital divide.A portfolio of a thousand songs can be bought for $1000 (roughly) on ITunes. But theoretically that portfolio can be reused by different members of the family on different devices for indefinitely thereby bringing the per unit cost of media down significantly.On the streaming model, it doesn’t work out that well — especially with models that place limitations on the number or type of devices (Zune Marketplace, Rhapsody, etc).Also because the streaming model is for a monthly fee, over the long run for a core set of songs works out much cheaper. I doubt it if most people listen to more than 2000 songs on a regular basis, and for keeping up with the latest music, you have services (for free) all over the net, and of course, conventional radio.
Streaming music is free for all intents and purposes nowIts the new radio
But it still isn’t a great experience. Walking home the other day I pulled out my iPhone 3G, and opened the Pandora app. Its hard to listen to music when it stops in the middle of the song a couple times to catch up. I also tried last.fm and it was even more annoying. And I had 5 bars on 3G mode. Seems like the network is the weak link….
Yup I can relate to that. We are bringing streaming video to mobile here in the states and it’s clear that even 3G isn’t enough. So we have to fall back on Wifi points. But you know what? Most mobile users have already been trained to look for a Wifi hotspot 1st and then to their network 2nd if they have a bandwidth intensive application. So wifi-only may be feasible in the early stages as long as there’s a biz model. Which means advertising. Which means the user doesn’t stay in control. Which is why we like streaming rather than VoD.
Totally. Streaming on mobile isn’t prime time yet but streaming on a wired connection is now. Mobile will get there
I fully agree. That is what will happen. The AGE OF ACCESS is coming. All the user need and want to is access to content.
And for all the disbelief in the technology – broadband or hardware – Moores law is still up to date.All the problems are solved in 18 months… don`t worry.
Streaming is no more on-demand than any other form of net access. (Yes, I know that torrents come out of order, but there’s no law that says that all downloading has to work that way. In fact, downloading used to be in-order, just like streaming. And, some streaming is already “near real-time”, using some time-limited out-of-order&p2p to help with bandwidth issues.) The user-experience difference is that with streaming, you can be disconnected from content that you “had”.From a technical point of view, streaming reduces the amount of storage required by requiring additional bandwidth for subsequent playback. It’s unclear that the bandwidth is cheaper than the storage. And, downloading can use bandwidth more efficiently. (A decent music service can predict what you’re about to discover and send it to you in advance when there’s extra bandwidth available, aka “the night before”.) It is clear that any bandwidth or service glitches make stream-dependent device less useful than those with storage.In the end, the dominant device will be a hybrid. It will stream when necessary, but store for subsequent playback. It will try to reduce “when necessary” by speculatively downloading what it thinks that you’re going to want “soon” whenever there’s otherwise unused bandwidth.It’s probably time to revisit the “how much would it cost to store all of the recorded music” calculations.
Speaking of using p2p for streaming – http://www2007.org/papers/p… – specifically addresses fast-startup for playback before the whole song/movie is downloaded.
I think there’s another huge advantage of streaming that you didn’t mention: instant gratification. Your audio or video appears far more quickly after you ask for it. This is much more fun. It also means that it’s a lot more feasible to try out an audio to see whether it’s something you’re interested in; the overhead (in terms of your own time) is so much smaller.
Totally true. And its a lot easier to apply social media best practices with streaming vs file based music
We all love to click and view video immediately. This is a great user experience and in many cases this works as a business model for content owners/distributors. But there are use cases where this will not be the best user experience nor a workable business model.To faciliate the delivery of free (ad supported) longer form, high quality video, P2P assisted delivery (interoperated with a CDN) will work for consumer and content owner alike. For consumers, this delivery will enable time shifted consumption (take the video with you on your next cross country flight) or subscription models that free you from seeking out the content (have the video automatically delivered via an RSS feed to your computer). For content owners, the economics are compelling. Even as traditional CDN prices drop, the economics of the current online video business model simply do not work. There is lots of talk about online video revenues but does anyone talk about online video profits?
Look at what gary vaynerchuck is doing with pleasedress.me to see how building a large audience can be monetized in many ways
Then we will need new ways of selling permanent access to media for consumers who just have to have a feeling of possession. In the post-file world, we’re really talking about “right-to-access” contracts.