Streaming Video

After seeing several red envelopes at a friends house the other day I started to wonder what Netflix‘s cost structure would look like if streaming video replaced sending DVDs by mail.   Mindful of Andrew Tanenbaum’s adage about never underestimating the bandwidth of a station wagon full of tapes hurtling down the highway, I thought it might make sense to do a cost model and see if streaming DVDs would be as cost effective as shipping them. This is a very simple model that does not take into account several crucial factors such as the First Sale doctrine, licensing for streaming,  partnering with studios instead of sourcing DVDs etc.  Leaving those aside and focusing on the technological aspect of the cost modeling is still quite illuminating.

The model is based on the NETFLIX INC (NFLX) 10-Q filed 7/31/2009. The results were quite surprising.

The article was improved thanks to contributions by Ben Black, Randy Epstein and Alex Pilosov


3 Responses to Streaming Video

  1. Director of Pump says:

    No one else has posted yet about this, so I’ll jump in.

    Yes, there are certainly potential savings by caching the content on a users individual DVR at home. One idea that has been toyed around with, is using multicast distribution. However multicast alone, has no inherent data retention. It would require a video provider to distribute out “hot content” to DVR’s ahead of time. Those who watched it, great. Those who didn’t? Wasted bits.

    Another thought would be to augment the service provider network to incorporate solid-state-disks in aggregation devices (DSLAMs, etc). However, such vendors in this space work poorly when it comes to layer 3, so forget about getting applications to run properly.

    The reason I bring this up is because someone always brings up multicast in a thread like this. The reality is multicast is just not well deployed in eyeball networks. If it is deployed (for developing broadband networks), its almost certainly locked out by the provider so that they can control any such multicast distribution (for good reason, as multicast state contributes to your scaling). In the end, the service provider holds all the keys here and its probably easier for the video provider to just unicast it out anyways.

  2. Owen says:

    I have read elsewhere that the cost of shipping physical disks is now higher than the cost of streaming movies, without any specific numbers attached. A few weeks ago I remember making the comment that there’s a good reason why they’re called “netflix” and not “mailflix”… and that it demonstrates more future sight than a lot of other companies. They bootstrapped via the mail to build a significant customer base that is now delivering the same revenue while their costs are only dropping. Any competitor is going to have to be a LOT better/cheaper to get any traction. On the other hand, licensing costs can’t be discounted and there is always a danger of retarded regulation that draws up completely different rules because something is “on the internet”. All that aside, I definitely like the streaming netflix service, it’s great with an internet->tv hookup.

  3. Jack says:

    Thanks for investigating this, very interesting post. It definitely got me thinking. It got me thinking so much, I had to make a blog comment that has sub-headings.

    Don’t forget HD –
    I’ve quickly become addicted to HD since upgrading the TV – I don’t have a blu-ray player (yet?) and I’m not sure if I want one, but if I did, clearly that would tilt the scale back towards shipping again, since it obviously is the same size and weight as a DVD.

    Back in reality, though, Netflix’s “Watch Instantly” is not sending 6 GB per movie – I don’t know what it actually sends, but it’s definitely more compressed than a DVD and I am pretty sure it even skips frames (maybe that’s just my stupid macbook’s fault). So it tilts back in favor of the internet, there.

    What I REALLY want, though, is real streaming HD, and in my experiences so far, it only looks good when I do that through the cable company. Netflix (and Hulu) look just OK, which is great if I just want to know what happens, but unsatisfying when I really want to take in the images.

    Peak-to-average Ratio –
    I’d be surprised if netflix usage fits a 1.7 for peak-to-average ratio. Maybe on the weekends, but for 5 days a week, for a service that I assume is USA only, the vast majority of people in our meager four time zones (with most people split between two of them, especially most people with adequate broadband) have a pretty narrow window for movie watching. Maybe I should just defer to the experts on that one.

    Multicast –
    I haven’t thought about Multicast in ages. I wonder if they could make it worthwhile by only allowing movie downloads to start at some fixed interval. That would definitely annoy people, but if they could get significant savings, even to the edges/peers of one backbone, maybe they could make it a billable option. That might be too complicated to explain for the average person, but perhaps not too complicated to live with.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: