Remember the “turbo button” idea? An Internet user could push it and get a burst of additional speed. Don’t sprain your fingers pushing at it, because it’s D..E..A..D. Remember the notion of a gaming-themed Internet service with ultra-low latency? It just received what’s likely a mortal blow, according to a Light Reading story. There are two ways you can read these news tidbits. First, any attempt to differentiate telecom connection services is doomed to failure, which means that for telcos it’s either rise up the value stack toward the delivered experiences, or get rescued in some way by governments. Second, any attempt to differentiate telecom connection services has to be tied to some very carefully crafted new experience, one that values something that’s not particularly valuable today.
The reason this is getting a lot of attention now is that there’s still a large community of earnest tech types who want to find value in 5G Standalone and network slicing. What good does it do to offer QoS specificity in slices if you’re not allowed to actually offer it at the retail level, or charge for it? But if retail priority services, even for gamers, isn’t viable, then what more could be said?
The Light Reading story recounts the experience of Cox Communications, who offered “Elite Gamer” services with up to 32% better game-server connection, then killed it because demand didn’t meet expectations. There have been a number of initiatives aimed at gamers, and my telco and cable company contacts tell me that there have been roughly three times the number of trials that didn’t result in deployment as there have been gamer-specific deployments.
Part of the problem with this sort of service is the net neutrality rules I’ve been blogging about recently. Regulators have been wrestling with the question of what users or OTTs should be able to pay for in the way of special handling. Originally, going back to the 1990s, the thought was that OTTs should not be able to pay for special handling, on the grounds that allowing that would favor giant incumbents and discourage innovation. That was broadened to include any form of paid prioritization, on the grounds that if consumers could pay for special handling then OTTs with deep pockets could somehow offer incentives to offset the costs. Now, at least in some geographies, we’re on-and-off-again with regard to neutrality overall.
OK, neutrality may be messing things up, but there are a lot of challenges with respect to enhanced connection services that extend beyond this neutrality issue. One big one is the question of how you assure what you’re charging for given the bill-and-keep model of Internet service. Some regulatory specialists have suggested to me that this is the real problem regulators have to face, because if you combine paid prioritization with bill and keep, you could end up with a service model that favors giants who can keep traffic on their own network longer, and thus control QoE better. On the other hand, telecom planners and some vendor strategists say that without some form of paid prioritization, things like the metaverse and edge computing may be difficult to advance, because the network might not deliver the potential benefits.
Business services, of course, should not be associated with neutrality regulations, but there are also challenges associated with premium handling of business services. Obviously, business Internet with premium handling could easily slide over into OTT-paid prioritization. Less obviously, since it’s difficult to know whether a “home broadband” is a “home office” broadband, it may well be that prohibitions on consumer paid prioritization would end up impacting business services. At the bottom of the “obviousness” scale is whether you could effectively prioritize a business service that crosses operator boundaries, given the general rules against paid prioritization in IP services.
The Light Reading story also talks about a possible (not confirmed) AT&T service that would offer mobile users priority in periods of heavy traffic. This makes many regulators I’ve chatted with highly uncomfortable; they see it as a way of raising ARPU by essentially making today’s service a kind of permanent underclass. Some within AT&T tell me they share that concern, which may be why the rumors of the service aren’t confirmed. Personally, I think that the heavy-traffic-priority point is riskier from a neutrality intervention perspective than a turbo button would be.
All of this validates the questions about the way that network services, and in particular the Internet model of network services, might impact future applications. Or how those applications might impact the Internet model, depending on how you rank the issues raised. Clearly, it makes little sense to talk about public edge hosting without latency control in the connection between the user site and the edge hosting site. The latency question also impacts the credibility of a global social metaverse, and perhaps the credibility of many of my “metaverse of things” applications. Finally, it impacts the value of things like 5G network slicing, because if you can’t charge for special handling what business value is there for operators to invest in slicing technology?
The flip side of this is the argument that if any paid form of enhanced network services is added, the result will be that best-efforts won’t be very good. In contrast, absent any paid enhanced service, these new things will be possible only by making best-efforts better, which benefits the open model of the Internet and favors users. That has been a very popular view up to last year, at which time the financial pressures on telcos seemed to make it unlikely that they’d invest in improvements to service quality without receiving more for their investment. This view is favored by the fact that while turbo-button services have fallen out of favor, some operators have been able to sell faster connections for more money to address improved handling. This, despite the fact that no current applications actually require the higher speeds.
The difficult question all this raises is whether we would have to make significant changes to the Internet business model, Internet infrastructure, and Internet regulations in order to advance the sort of applications everyone thinks will create the future of Internet services and worker IT investment. An even more difficult question relates to this one; even if we were to make changes in all those areas, would any of the new applications arise? In other words, should we worry about preparing the markets for something whose business case cannot be made?
The gaming service concept was and is important; we need to find applications for new services that push the capability limits of current services but aren’t so totally off-the-wall that they can’t be launched without massive speculative investments in new stuff. That’s the only practical way to explore what new services will look like.
I think that is what will have to happen in order to set realistic expectations for the value of the applications, and also realistic improvements in networking that could increase that value. However, because current service limitations will constrain these experiments, I think they’ll take longer to conduct. If we had a fertile field of stuff out there waiting to be exploited, we’d expect exploitation to explode. Absent that, it will blunder about until it finds an optimum path, and we’ll just have to wait to see how well that path suits our vision of what the future should be. But for those 5G fans out there, I have to tell you that 5G can only win, or even be relevant, if an application experiment that validates some aspects of it emerges and is successful. If you want a good 5G outcome, stop blathering about features-in-a-vacuum and start thinking about application value. That’s an important lesson elsewhere too. Edge isn’t value. APIs aren’t value. The two are ways of exploiting value, but for either of them or for 5G to work, the value has to be created, and only applications can do that.
With this, I’ll leave topics related to neutrality and the Internet business model, as the two relate to the future of telecom services…unless something new comes up, of course.
