The more I think of yesterday’s comments by the FCC’s Tom Wheeler about paid prioritization and fast lanes, the more I see the fingerprints of Apple. Remember the rumored Apple/Comcast discussions a couple months back about a separate pipe to the consumer?
Wheeler seems to be trying to thread a needle through it all, simultaneously promising no slowing or blocking of anything the consumer has bought but leaving open a door for something else. I do think that something else could be *additional* bandwidth paid for on the content that supplements what the consumer buys.
Is that a distinction without a difference? Let’s step back from the knee jerk reaction that one man’s fast lane is another man’s slow lane for a second engage in a thought experiment. Let’s go further down the path of what a reader of my Apple fast lane article described as a VNO type of system.
Suppose the FCC allowed last mile operators to sell to the content/cloud/CDN world quantities of supplemental bandwidth to the consumer, in excess of whatever broadband plan the consumer is paying for. The extra capacity could be over any access method – wireless, fiber, dsl, cable – and could vary in size for each. They would then become an Access Virtual Network Operator, or AVNO, a classification the FCC could then regulate in some form.
Thus, someone like Apple could buy its own and use it to deliver some or all traffic for its own closed ecosystem, but someone like Amazon or Akamai could turn around and sell access to theirs to their cloud or content community in bite sized quantities. You could think of it as just an optional extension of the CDN infrastructure we have today. The big guys roll their own, the small guys buy from aggregators only what is needed while the pricing leverage is handled through the larger ecosystem.
There could be any number of such AVNOs, so you might have an ‘up to 50Mbps’ data plan, but the dsl/cable/fiber modem could also be delivering the consumer a number (let’s say 3) of ‘up to 10Mbps’ in AVNO pipes with specific content. The consumer might not use any services that could access those pipes, or those he does might be explicitly paid for like Netflix or via advertising on the other end.
Put it all together:
- The consumer gets more bandwidth for his buck. He still gets the pipe he paid for, and some extra that kicks in for some traffic.
- The last mile gets another revenue source to pay for an effectively wider bandwidth to the consumer, thus encouraging investment without fear of OTT. But they do have to build out the infrastructure for *more* bandwidth in order to get paid.
- Does the smaller content provider get squeezed out? Well, if he is now able to tap an ecosystem of AVNO bandwidth in the amount he needs without having to negotiate with the last mile, the hill doesn’t look as steep as it would if everyone had to do a secret deal with the last mile. In fact, it doesn’t look any different than paying for CDN services does today on the other side of the network.
- Do large content operators have better access than small? They have that today everywhere but the last mile anyway, as they can buy better transit, CDN, and data center services in the first place. Size always has advantages and disadvantages. The key has to be that the path for the small agile provider mustn’t be artificially choked off. With a VNO-like structure, their market power can be aggregated to help compete.
- Does am AVNO operator have a conflict of interest between his own traffic and that of his customers? No more so than the last mile does today. And if the AVNO owner is also held to the standards of no prioritization/slowing of traffic, then not at all.
- Does the last mile have an incentive to squeeze the consumer’s original bandwidth itself in order to maximize its revenue from content? This is surely the biggest sticking point, but perhaps it’s one that actually can be regulated. It’s pretty straightforward to monitor raw congestion and pricing once the rest is worked out.
- Does this solve the other peering/transit layer issues? Not even close, but you can’t have everything.
- Isn’t this overly complicated? Very possibly. Why bother with all this if we just want a 1Gbps pipe to the home we can use as we please. But if it doesn’t work economically, then no AVNOs get created right?
All the FCC has to do is regulate the conditions surrounding VNO creation, enforce some form of local strong network neutrality on each VNO and the core last miles, and keep them honest about congestion and pricing for the core broadband connection.
Does that work? I don’t know, I’m just trying to follow the thread to see where the network neutrality discussion might wind up. I don’t think it’s helpful to turn this into a good versus evil, take no prisoners kind of thing. We need to find a path to an ecosystem that works for everyone, encouraging innovation in content and applications while ensuring that network investments are rewarded.
If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!
Categories: Government Regulations · Internet Traffic
Join Our LinkedIn Group
Rob, clearly you gave more thought to my comment on that Apple discussion. 😉 Only difference is I called it a Fixed Line Virtual Network Operator (FLiVNO) and you’re calling it a Access Virtual Network Operator.
As noted in my comment on the Apple discussion, I have a tough time believing ISPs like ATT/VZ/CMCSA/etc. will embrace this.
This isn’t merely resale, this is network unbundling. I don’t know how far back your telecom history goes, but when the 96 Telecom Act passed, one of the provisions (I believe Section 251 or Section 252) permitted Total Service Resale which was extremely unprofitable for carriers b/c it was top down — that is, Total Service (Retail) Revenue less the cost of servicing the customers leaving roughly 5-10% gross margins for reselling CLECs. CLECs and IXCs hated it. AT&T (then a CLEC) came up with this concept of rebundling. Essentially, because the Telecom Act permitted “unbundling”, ATT said, we should be able to unbundle the entire ILEC network and “theoretically” put it back together.
What was the difference between “Rebundling and Total Service Resale?” The rebundled pricing was based on forward looking costs using a bottom up or Total Element Long Run Incremental Cost (or TELRIC). Consequently, it was cheaper than resale. More importantly, rebundling also permitted the the CLECs to keep (or not charge) for switched access charges on long distance calls. Under resale the ILEC got the access charges.
As I said, I don’t see ISPs (ILECs) embracing this model.
I forgot to add that if past performance is indicative of future behavior, expect massive legal battles from the ISP community. Just as Bell Atlantic, NYNEX, BellSouth, PacBell, Ameritech, USWest and SBC fought the ATT rebundling proposal and tied up the administrative courts for years, I’d expect the ISPs to fight this no less agressively.
How will the ISPs fight it — among other arguments, 5th Amendment Takings Clause violation.
Let’s not get ahead of ourselve though. First, ISPs won’t agree to it. This will force the FCC to threaten (or impose) Title II reclassification. Then after years of wrangling on Title II reclassification, the FCC will order the ISPs to provide this FLiVNO solution to other players. The ISPs will then argue that FLiVNO is a 5th Amendment “takings” violation — essentially, the governement taking private property without just compensation.
You may say, well, fine Apple will pay for a 100g IP highway on the ISPs’ networks. But ISPs will dump into the cost of those superhighway everything but the kitchen sink. They’ll say we have to be compensated, not just for the physical network, but also for the cost of our Operations Support Systems (OSS), for example, our billing and customer service systems, which were scaled for a much larger subscriber base than we’ll have if you force us to let Google, Apple, Disney and Amazon take highways on our network.
As much as I’d like to see this happen, it’s hard to imagine it will. I imagine ISPs will tie this up in the courts for years.
The only solution which will foster investments in new technologies to compete against the embedded ISPs, is cost causation.
Maintain net neutrality (this is key), dare the ISPs to raise subscriber rates (through a usage-sensitive rate structure) and bring in the DOJ’s antitrust department to police the ISPs’ behavior. When that happens, investment dollars will flow for new technologies and spawn a new breed of Competitive Local ISPs (CLISPs).
The DOJ rarely, if ever, uses its antitrust department to actually ensure that monopolist/duopolists are not abusing their power. Instead we see the DOJ’s antitrust department rear its head on merger reviews and long after anti-trust violations have been recorded. Put fixed line ISP duopolists on notice today that the DOJ’s antitrust department is carefully monitoring their behavior. Force them to provide cost support for raising subscriber rates (which they’ll easilly do b/c ILECs are masters at fabricating, i mean producing, cost studies. That’s ok though, b/c this will constrain their desire to raise rates. In the meantime, investment in new competing ISP technologies will surface and the CLISP market will be born.
There you have it: The next 15yrs of the NA communications network evolution. Of course, if the DOJ’s antitrust department had just said no in the first place to NYNEX/Bell Atlantic merger (announced 5 months after the passage of the ’96 Telecom Act), we wouldn’t be having this discussion today.
While it’s an unformed idea, my thought is not to force the VNO model and regulate its pricing. Rather I’d force strong net neutrality on the base service while allowing the VNO fast lane model on top with limited intervention. Let the ISPs and content figure out a VNO model that works for extra bandwidth while keeping the main channel the customer pays for as Wheeler describes it.
Rob, why do you think ISPs would negotiate in good faith with VNO wannabes whose sole purpose is to steal higher margin sticky consumer customers?
Under the VNO model,
(1) the VNO (e.g., Apple) pays for a superhighway of with specificed throughput on the ISP’s (e.g., CMCSA) network.
(2) the VNO packages unique content (e.g., ITunes, concerts, movies, cable TV content or sporting events) internet access and maybe their own form of cable TV for its customers.
(3) subscribers pay the VNO for the unique content AND internet access NOT Comcast/VT/T
(4) ISPs only source of revenue is from the VNO.
To repeat my lead question, why would an ISP in a duopolistic market forgo higher consumer margins and revenue and instead, voluntarily, sell a lower margin wholesale service to customers (i.e., VNOs) that want to steal my market?
I’m not saying this is bad for consumers. In fact, I believe just the opposite. Deep-pocketed Apple, Google or Amazon could probably take a fair amount of market share from ISPs quickly under a VNO model. At some point these players would probably look to move off the VNO model and construct their own network. (Again, this is great for consumers.) At the same time, however, this leaves a tremendous amount of stranded investment on the ISPs (i.e., CMCSA/VZ/T).
Unless I’m missing something I just don’t see what’s in it for the ISP.
CCL,
I can think of few differences which could help last mile providers think differently about VNO partnerships:
1) MSOs and wireless providers are more suited to revenue-share models than wireline ILEC/RBOCs of yesteryear. AT&T provided all of the value and received all of the revenue on the wireline phone network, but MSOs and programmers have always split the pie, as have wireless carriers and device manufacturers (and app developers, even before smartphones).
2) Apple proved this model can be healthy for a carrier when they were exclusive with AT&T Wireless on the iPhone. The carrier sacrificed control in favor of a defined revenue opportunity and it paid off.
3) Apple also proved revenue sharing can be viable for a content distributor (iTunes/App Store) and content creators.
4) Apple has given up ground of late in the smartphone wars and needs a big, game-changing win. As such, they may agree to terms which the early mover last mile network would find favorable.
5. Last mile providers will bet on being able to sell Apple a VNO fast lane while *retaining* their core access revenue, so VNO revenue will initially be additive. And enough residual value may remain in the “legacy” video products that AppleTV subscriptions come to be viewed as premium add-ons (like HBO or Netflix) and not necessarily a substitute.
6. For last mile providers like Comcast/NBCU who also own content and advertising businesses, the VNO model also potentially provides a new distribution channel which has a built-in home-field advantage.
Anon, these are all excellent points. I agree a precedent of successful revenue sharing does exist in the pairing relationships you provide.
In each of the arrangements you cite the combination created a greater whole because both sides complemented each by contributing something the other didn’t have. Moreover, the brands in the revenue sharing arrangements above enjoy equal prominence.
I just wonder if the Fixed Line Virtual Network Operator (FLiVNO) truly complements the ISP the same way the above-referenced pairings do. The FliVNO purchases a superhighway on the ISP’s network for the sole purpose of competing against them. I suppose the ISP (e.g., Comcast) can become the Intel in the relationship. For example, “Apple I-Flash brought to you by Comcast [insert pithy brand name]” where Comcast is the “Intel Inside.”
We’ll see how (and if) these unique arrangements unfold.
Yes I surely did think more about it, and I do think it’s time we stopped treating the wireless side differently.
As for unbundling, I do see the parallels. But just because it was done poorly once doesn’t mean something like it can’t work. The CLECs had no leverage other than regulation and thus were at the mercy of changes in the rules, whereas the content side has substantially more leverage since they have the content and a better relationship with the consumer to boot.
Rob, the concept of a Fixed Line Virtual Network Operator (FLiVNO) is not entirely new. During the European Commission’s (EC) local loop unbundling (LLU) proceedings (2000-2003) the concept of “Bitstream Access” was proposed. Bitstream Access was similar, but went much further than ATT’s rebundling or UNE-P (for unbundled network elements platform) proposal.
As you might imagine, Bitstream Access was hotly contested. In the end the EC’s final LLU rules did not mandate Bitstream Access. If memory serves, regulators were not considering imposing Bitstream Access on cable operators. (In fact, cable modem access was just introduced so the cable companies’ involvement was limited to carefully watching the LLU proceedings and protecting against any and all cable unbundling obligations.)
Only a small portion of the Other Local Operators (OLOs), Europe’s less pithy name for CLECs, focused on pushing the EC to force the EU’s ILECs to make Bit stream Access available along with the more traditional LLU solutions, like dry copper. The most vocal EU OLOs were facilities-based (e.g., COLT) and they primarily focused on dry copper for DSL quality loops.
One of the arguments against Bitstream Access posited by some EU ILECs was asymmetric regulation between cable and telco. In other words, ILECs argued that you can’t force us to provide Bitstream access when cable doesn’t have to. However, back then it wasn’t quite as compelling an argument b/c Cable ISPs were in their infancy.
A 2010 white paper entitled “Access regulation, competition, and broadband penetration:
An international study” discusses the idea of Bitstream Access, among many other ideas, as a possible solution for stimulating more broadband competition
http://202.114.89.42/resource/pdf/5451.pdf