Hacker News Reader

Comment by cluda01 | Hacker News Reader

Hopefully they should be able to use these cost savings to purchase more content and get it into their system. Does anyone have any informed speculation about how much they could potentially stand to save? I'm unfamiliar with how peering agreements work and or the cost structure.

View on HN

It's interesting from a bit of a game theory standpoint. The optimum outcome might be achieved if Netflix hands over the appliance and the ISP installs and keeps it running at no cost to either party, except for maintenance costs (hardware replacement, etc.) which Netflix bears. Both parties save.

If Netflix wanted to get clever, they could calculate the cost an ISP would likely pay for the bandwidth their users consume, and charge the ISP say 40% of this amount to host the appliance. Netflix saves 100% of its third party CDN bandwidth costs for that ISP and the ISP saves 60% of their peering bandwidth costs.

If the ISP wanted to get clever, they could calculate the bandwidth that Netflix is paying a third party CDN for the traffic consumed, and offer to host the appliance for 40% of this amount. The ISP saves 100% of its peering bandwidth costs for Netflix traffic and Netflix saves 60% of their CDN cost.

What's likely to happen is a bit of cat and mouse over the costs. The ISP peering is likely a lot cheaper than Netflix's CDN costs, so this could work out as:

    Netflix CDN cost
  - ISP peering cost
    ----------------
  = Cost to Netflix

This is all pure speculation with made-up percentages, of course.

comcast is actually charging people more than transit rates to get access to their network -- which is why (except for level3, who has a special deal) video on comcast sucks even with large comcast pipes.

Actually, no, Comcast's paid peering costs are significantly cheaper than all but the lowest quality transit. In other words, if you just have Cogent, you get what you pay for.

That may be true at the 10G level (which I didn't price), but isn't true (in my experience) at 1G. Plus, it's a lot more likely that you'd want 10G of transit and 1G of comcast than 1G transit/10G comcast, so per-G, your decent quality (Level(3), nLayer, etc.) transit is going to be cheaper. Comcast also seems to negotiate less, since they're effectively a monopoly on routes to themselves, vs. transit.

(Obviously the right thing to do is grow to needing 10G of comcast and 50-100G of transit...)

Fair enough. The last time I was privy to this, it was more than a year ago. I didn't expect things to change that much :-).

Netflix already said the box is free (as long as you're big enough to actually need it).

I think the content question will never be solved as long as

a) Cable companies put the squeeze on content owners to give them exclusive rights

b) Netflix religiously clings on to both commercial-free and buffet-style pricing

It's tricky. You have a movie studio that pumps out AAA features that cost upwards of $200 Million to make, then Netflix comes along and says "hey, you mind taking this really low price for your content, hopefully in perpetuity?" What would your response be? So, it may be awhile before Netflix gets the big-budget blockbusters quickly. That is, unless, their subscriber base goes over 50 Million, then by ALL means game on.

I think that, in the interim, a dynamite move for Netflix would be to acquire a music streaming company like Pandora or Spotify, then a streaming game company like Galkai or OnLive. Push it out on a set top box, a la the OnLive console, charge $19.99/month with the opportunity to rent premium newly-released blockbuster movies, hot records, and AAA games on the day of release for the same prices that redbox or gamefly charges to rent them per day.

Reply on news.ycombinator.com