Dumb, Telecom Industry Backed ‘Network Fees’ Drive Twitch Out Of Korea

from the please-pay-me-extra-for-no-coherent-reason dept

Twitch has announced that the company is shutting down in Korea after regulators there imposed a ridiculous new regulatory framework that drove the company’s operational costs through the roof.

Basically: Korean telecoms convinced gullible regulators to pass a new regulatory framework wherein edge providers and content companies are forced to pay telecoms additional fees just to have their traffic successfully reach its destination (consumers). It’s in addition to bandwidth costs, and, as we’ve pointed out for a while, it’s a dumb cash grab and the latest extension of the longstanding net neutrality wars.

In a blog post, Twitch CEO Dan Clancy explains that with the addition of these new mandated network fees, the company has found it impossible to turn a profit in Korea and has been forced to close up shop:

“Ultimately, the cost to operate Twitch in Korea is prohibitively expensive and we have spent significant effort working to reduce these costs so that we could find a way for the Twitch business to remain in Korea. First, we experimented with a peer-to-peer model for source quality. Then, we adjusted source quality to a maximum of 720p. While we have lowered costs from these efforts, our network fees in Korea are still 10 times more expensive than in most other countries. Twitch has been operating in Korea at a significant loss, and unfortunately there is no pathway forward for our business to run more sustainably in that country.”

For the better part of the last few decades, global telecom giants have been trying to “double dip.” As in, not only do they charge businesses and consumers an arm and a leg for (often terrible) broadband, but they’ve been trying very hard to institute new regulatory frameworks where consumer and content companies alike pay telecoms huge sums of additional money for no coherent reason.

This idea that big ISPs are inherently owed a cut of the revenues of services traveling over their networks is what launched the net neutrality wars around the world several decades ago, when AT&T insisted Google “wouldn’t ride our pipes for free.” It’s evolved in dumber and dumber ways ever since.

Under Korea’s model, edge providers (like Netflix) are forced to pay “network service fees” to ISPs. Basically, ISPs there have claimed that they’re inherently owed more money if a TV show on Netflix is super popular, claiming they should be compensated for extra bandwidth costs.

Of course, bandwidth provisioning doesn’t really work like that. ISPs are supposed to build networks that can handle any peak capacity spikes caused by normal consumer demand. The origins of those demands are irrelevant. Consumers and edge providers have already paid an arm and a leg for bandwidth, particularly if regional monopolization has driven down any incentive to compete on price.

All they’re really doing here is trying to offload network operations and maintenance costs to someone else. In this case: Korean game streamers or Netflix users.

Demanding that popular companies pay more to telecoms just for being popular is an inherently stupid idea, but it’s been dressed up by telecom lobbyists as serious adult policy under terms such as “sender pays” or sometimes “Sending Party Network Pays” (SPNP). I’ve been dumbfounded by how these proposals have been treated as serious policy.

The efforts always begin with false claims that companies like Google and Netflix are somehow “getting a free ride” on the internet, despite spending billions in bandwidth, CDNs, undersea cables, and cloud infrastructure. From there, they usually involve some flavor of false claim that this model will help expand broadband availability to those in need. But its only real function is to fatten telecoms’ purses.

Like many countries, Korean regulators largely favor just three big major ISPs, which then influence policy determinations to an extremely lopsided degree. This muted competition, combined with regulatory capture, plus the SPNP model, has driven up costs for Korean consumers (see Michael Nelson’s good 2021 piece on this, or this 2022 Techdirt story by Konstantinos Komaitis).

Now, regulatory capture is driving edge companies out of business, and driving up both bandwidth and content costs for consumers. It’s all predatory nonsense created by regulatory capture and corruption, and the telecom policy marionettes pushing the idea aren’t operating in good faith. They’ve hijacked regulators to implement systems that deliver telecoms billions of additional dollars for doing nothing.

And if you think this is only happening in Korea, it’s not. It’s been an ongoing debate in the EU, where telecoms have floated a direct tax on any company that accounts for more than 5 percent of a telco’s average peak traffic. Here in the States, FCC Commissioner Brendan Carr (R, AT&T) has been a big proponent of the idea for several years now.

Again, it’s basically telecoms trying to get paid twice for the same (often substandard) service. That it has been dressed up as a serious policy proposal is embarrassing. California’s net neutrality rules ban such practices, and if they’re worth anything, the FCC’s soon-to-be-restored rules will as well.

Filed Under: , , , , , , , , , ,
Companies: amazon, twitch

Rate this comment as insightful
Rate this comment as funny
You have rated this comment as insightful
You have rated this comment as funny
Flag this comment as abusive/trolling/spam
You have flagged this comment
The first word has already been claimed
The last word has already been claimed
Insightful Lightbulb icon Funny Laughing icon Abusive/trolling/spam Flag icon Insightful badge Lightbulb icon Funny badge Laughing icon Comments icon

Comments on “Dumb, Telecom Industry Backed ‘Network Fees’ Drive Twitch Out Of Korea”

Subscribe: RSS Leave a comment
25 Comments
Ehud Gavron (profile) says:

Re: Re: Refunds for downtime

Probably. Big purchasers like that will often have a service level agreement (SLA) that specifies refunds for down time.

One could only wish.

Typical SLAs have specific limits and thresholds:
1. 99.9% (yes, that’s e 9’s, not the 5 9’s they said in the brochure).
If you do the math, 99.999% uptime means no more than 5 minutes downtime a year. 99.99% uptime means no more than 50 minutes downtime in a year. 99.9 – 500 minutes – 1 business day.

  1. Pro rata CREDIT (not refund and there’s a difference explained below) for time from “OUTAGE START + REPORTED” to “OUTAGE CLAIMED RESOLVED.” As you can imagine, on a timeline REPORTED to CLAIMED RESOLVED is significantly less than “START to OUTAGE RESOLVED.”
  2. A pro-rata credit on a 99.99% SLA for a one hour outage is 10 minutes. That’s 1/6th of an hour, 1/144 of a day, and less than 0.02% of a month. If your monthly bill is $5,000 (going rate for 10G transit) then you can get that whopping $10 sometime soon.
  3. Refunds are required to be issued in the next billing cycle unless the T&Cs say otherwise. You did read them, right? Because talking of SLAs without mentioning the 9’s or the pro-rata suggests NOT. CREDITS can be issued anytime up until contract termination and are sometimes used to offset early termination liability (ETL). You can imagine that your 60 minutes of downtime when allowed 50 minutes of downtime means you can term early by 10 minutes with no penalty.

In my former company we offered a 100.00% SLA. It required 3 physical (layer 1) circuits, 3 carriers (layer 2), and 3 IP networks (layer 3) and routing protocols (layer 4) to ensure full bidirectional connectivity. It was expensive, but nobody ever called to say “We’re done!!” or “For every hour my chiropractic office is down I’m losing $100,000” or “babies will die.” I’ve not seen that offered elsewhere.

In various capacities and roles I’ve negotiated contracts and the key elements are:
1. SLA percentage of uptime
2. Removal of percentile scoring
3. Immediate refund or credit of a higher-tier amount, not the pro-rata
4. Documentation. I don’t want to hear “try it now” or “we think we figured it out.” Figure it out and fix it and then let’s talk.

Ehud

Ehud Gavron (profile) says:

This didn't start with AT&T and Google

…and it didn’t start being called Net Neutrality.

It was a mid 1990s issue where ONE COMPANY decided that they would DENY peering to ANOTHER COMPANY and require/force that company to BUY TRANSIT.

Transit and peering are clear concepts to industry people, but to laypeople they are not. The words “net neutrality” became common over the following two decades (yeah, that long) to make it easier to understand.

The bad guys: Level3. Hands down. History and facts are a reality that we live with.

E

So, if you read this far, surely you’d just dying to know what these words mean.

TRANSIT: If you pay me, then I will take your bits and bytes and send them to the next point along the way to its destination. I will accept bits and bytes for your IP addresses and send them to you. You will pay for this.

PEERING: You have a lot of traffic that MY customers want. You receive a lot of traffic that YOUR customers want. Also my customers want to send you traffic, and your customers want to send mine traffic. How about this… so long as it’s a fairly even trade of bandwidth, we’ll just connect this’a’here router port to that’a’there router port and we’ll do some chingering with BGP[4] and ASNs and routing policies, and route registries and routing authorities (which are NOT authoritative) and we’ll just make it all work hunky’dory’k? No need to write a check. We’re just happy to work together.

NET NEUTRALITY: It’s whatever the legislative or regulatory bodies choose to invent it to be at the time they so choose. It’s somewhere less expensive than transit, and somewhere more expensive for peering, and at the end of it all we the people pay more for the same thing.

ISPs – provide Internet service and nothing else.
TELCOs – provide telecommunication services (originally all TDM)
CABLECOs – provide services using their own cable systems (now DOCSIS)

Hope that helped.

Mamba (profile) says:

Re:

Your whole argument seems suspect to me. In the mid 90’s Kewit (soon to spin of their telecom subsidiary as Level 3) didn’t offer IP products. Their infrastructure was all ATM at the time so an interconnection would have been by definition transit. They didn’t announce until 98 that they would build out an IP network, and I’d be surprised if they had any fiber in use until 99.

Ehud Gavron (profile) says:

Re: Re: History is facts, not suspicions.

Your whole argument seems suspect to me.

Your suspicions do not change facts.

In the mid 90’s Kewit (soon to spin of their telecom subsidiary as Level 3) didn’t offer IP products. Their infrastructure was all ATM at the time so an interconnection would have been by definition transit.

That makes no sense whatsoever. Transit of IP traffic IS AN IP PRODUCT that they CLEARLY WERE SELLING.

They didn’t announce until 98 that they would build out an IP network,

Their public relations strategies and timing of announcements is not something of any relevance.

and I’d be surprised if they had any fiber in use until 99.

Your surprise or lack of education in the history is not something of relevance.

All these things you said are entirely irrelevant except for the 1998 part. Whether layer 1 delivery was fiber or copper is irrelevant. If they used ATM or MPLS or whatever because TDM was just how things were… and… like the rest of your comment… irrelevant.

Small-time Colorado company started a thing which has snowballed over more than two decades. That’s a real thing.

Next why don’t you tell us all how you’re “surprised” and “suspect” that Canter and Siegel didn’t create spam. Or that Elon Musk didn’t get bailed out by his PayPal cohorts to the tune of Tesla’s current success. Really if you have nothing to say “I suspect” and “I’m surprised” could be better spent saying nothing.

The commercial Internet has been around for a long time – 30 years if you would like a timeframe. During that time there have been various shifts in positions of thought. (Nothing as exciting as “suspicious surprises”, but ok.)

There is no more conceptualization of “THE backbone”, “Tier-1 ISPs” etc. The same generalized concepts that went away means the difference between transit and peering and “network neutrality” is just a matter of definition… and contracts… and extortion.

You want to defend Kewit or L3? Feel free. Their extortionist behavior started this… and it won’t ever end.

Mamba (profile) says:

Re: Re: Re:

Jesus fuck that’s incoherent. You seem to have some confusion.

Transit of IP traffic IS AN IP PRODUCT that they CLEARLY WERE SELLING.

Clearly was selling? Why do you say they were clearly selling IP product in the mid 90s?

Here’s what Kewit has to say about the timeline

. In 1993, MFS was spun off as a separate company and taken public, with Crowe appointed as its leader. The future of the company was changed significantly a short time later, after Scott attended an exclusive, biennial gathering hosted by Warren Buffett. Bill Gates, the cofounder of Microsoft Corporation, gave a speech in front of the 50 guests, stressing the revolutionary nature of the Internet and how his company was responding to the advent of the Internet age. Gates also remarked that the new technological advance could render the business of traditional telephone companies obsolete, a comment that pricked Scott’s ears. “Yes, Bill’s talk is probably what got us thinking about the Internet,” he reflected in a July 22, 2002 interview with Telephony.

and

With Scott’s blessing, Crowe’s new venture was organized within Kiewit Diversified Group (KDG), which comprised all of Peter Kiewit Sons’ businesses and investments in non-construction areas. Crowe began building his start-up company in mid-1997, enlisting the help of 18 former MFS executives whom he recruited from WorldCom. His first objective, under the guidance of Scott, was to sell KDG’s non-telecommunications assets to provide the start-up capital for building an expensive and vast fiber-optic network. Crowe’s company was a start-up, but one with enormous financial resources behind it from the start, putting it in an entirely different league from other venture-backed, start-up enterprises.

https://www.referenceforbusiness.com/history2/19/Level-3-Communications-Inc.html

Here’s what the FCC says about the timeline:

Level 3 will offer a full range of communications services beginning in the third quarter of 1998.
https://transition.fcc.gov/enbanc/070998/level3.pdf

Ehud Gavron (profile) says:

Re: Re: Re:2 Jesus fuck that’s incoherent

If you want to go back to MFS and WCOM and then disclaim L3 was selling transit, you may want to look at original sources. Unfortunately history wasn’t being stored like it is now so it takes more than a click on the G icon to get there.

https://www.itu.int/en/ITU-D/Regional-Presence/AsiaPacific/SiteAssets/Pages/ITU-ASP-CoE-Training-on-/itu-asp-coe-te-session6_internetevolution.pdf

I love it that your heroes had a meeting in 1998 — a full five years after the MAEs and the CIX were functional and nobody thought any more about NSFnet but Bill Gates and Warren Buffet were amazed.

1998 wasn’t the breakout year of anything…. except for this… https://www.imdb.com/title/tt0142134/

History is ALWAYS re-written by the victors. If you’ve come along a generation too late you have nothing to trust, but as you said “Jesus fuck that’s incoherent.”

E

glenn says:

So then what are the fees that subscribers pay even for anymore if they don’t guarantee full Internet access and a complete round-trip for our traffic? I would feel like I’m paying a full fare to get on a bus, but I won’t be allowed to get off until another full fare is paid either by me again or by someone at my destination. That’s not business; that’s highway robbery… or just a protection racket. This is just government-approved and -supported crime. I’ll be cancelling my sub now.

LostInLoDOS (profile) says:

Reality

This one doesn’t make sense beyond the false assumption.
Yes, some companies abuse unmetered access. Intentionally.
But Korea has one of the strongest internet environments in the world.

This is upside down! To start with: many authors don’t understand, can’t comprehend, the level of data being transmitted. We’re not talking gigabytes, terabytes. We’re in the realm of petabytes. Or more. Even in countries as robust in deployment as Korea there’s a huge drain and clogging here.

But: double dipping isn’t the solution. I’m not sure what is, beyond metering. And that’s not a solution I like. But when you drive so much data banks and hospitals have delays during peak hours… there’s an issue.

Who should pay for the fix is a major question, one I don’t have an answer for. But let’s not pretend that these streaming services are not without cause.

Add Your Comment

Your email address will not be published. Required fields are marked *

Have a Techdirt Account? Sign in now. Want one? Register here

Comment Options:

Make this the or (get credits or sign in to see balance) what's this?

What's this?

Techdirt community members with Techdirt Credits can spotlight a comment as either the "First Word" or "Last Word" on a particular comment thread. Credits can be purchased at the Techdirt Insider Shop »

Follow Techdirt

Techdirt Daily Newsletter

Ctrl-Alt-Speech

A weekly news podcast from
Mike Masnick & Ben Whitelaw

Subscribe now to Ctrl-Alt-Speech »
Techdirt Deals
Techdirt Insider Discord
The latest chatter on the Techdirt Insider Discord channel...
Loading...