I'm a big believer in 4G. Not just because I can do certain *tasks* faster, but because it just *feels* fast. Some of that perception is from quicker connection and start-up times than 3G, some is from impressive headline numbers when I've run a speed-test, but a lot is from a sense of "potential", knowing it's there if I need it.
It transpires that something similar is true of fibre in the fixed world - especially in FTTP (fibre to the premise) guise, according from some proper consumer research done by friends at Diffraction Analysis. People like raw speed, are annoyed when they can't get it, and are prepared to upgrade further once they do. They might even use it to consume extra operator-enabled services as well.
One of the regular Twitter debates I have with colleague Martin Geddes is whether broadband "performance" is about specific, measurable, application outcomes or not ("QoS" to most of us). He links that to an argument for non-neutrality, as he contends that specific Internet application providers have different performance requirements and should be allowed to "trade" for them, given the limits of the network.
Conversely, I argue that a lot of the benefit and value from fast, open (and neutral) networks is not application-specific, but rather is intangible. It's the potential for both users and developers to do what they want today or in the future, without trading or risking potential competitive abuse or extra friction added by ISPs.
It's like having a powerful car but sticking (mostly) to the speed limit - you know the "shove" is there if you need it. That extra acceleration on the highway on-ramp doesn't get you to the destination faster - but it feels good. Some even pay a premium for unreliable or twitchy supercars, despite the chance of a bad (even catastrophic) "outcome".
We're human. Perceived performance - and flexibility - is often more important than measured performance. Doesn't matter if it's broadband, cars, fashion, headline megapixels in a camera or a million other areas of life. There is almost always some correlation between "More of X" and "Better outcomes", but even if it's not a perfect correlation, we don't care.
Purists myopically focused on "optimisation" often don't understand this, or other human emotions like agency/control, a sense of novelty, image and so forth. While these apply more to individuals than businesses, there other factors like privacy and security and agility too.
In other words, "just give us faster networks & then get out of the way. Occasional glitches are a price well-worth paying for freedom and permissionless connectivity".
It's true of 4G - most people upgrading are happier than with 3G, although obviously some of that is down to the device itself, or perhaps better coverage with 7/800MHz vs. 2100MHz on some networks. Many notice the initial "whoosh factor" - that it's just fast even if they're doing something that doesn't need it. Sports-car feeling, again. And over time, developers are exploiting that, even if they know that there will sometimes be issues they have to work around
And that's the thing - the vagaries of mobile networks (coverage, congestion and so on) have taught application and content providers to work around performance limitations and occasional failures. To expect them and plan for them - so they use cacheing, variable bitrates, UI interventions to warn users of problems, clever codecs and error-correction. There will always be a weakest link. It's not perfect, but it beats having the hassle, friction and perhaps commercial conflicts-of-interest involved in paid QoS. It's why "non-neutral" mobile business models won't succeed, even if the law allows.
But back to fibre and FTTx, and some hard data. I don't often reference other analysts and consultants. (Martin G is one that I collaborate on about voice/comms, and disagree with on networks). But when it comes to the fixed-broadband world and especially fibre networks, access business models and wholesale metro, I'll gladly defer to Benoit Felten, who's been covering that beat for years. He now runs his own research firm, Diffraction Analysis, originally out of France although he's now living in China.
Many telcos have been slow at rolling out fibre, often because they have been unconvinced that consumers really want it, would pay more for it, or might adopt additional services as well.
Diffraction has been doing some interesting work in collaboration with the FTTH Council Europe, looking at the real-world experience of consumers who have fibre broadband, and comparing it to ADSL. Benoit has now published research on Sweden (one of the most developed FTTx markets in Europe), and has been working on France and Portugal studies as well.
I've had a chance to have a look through the Sweden report, and it corroborates my views in various ways (although it doesn't tackle neutrality, per-se):
There's a lot more in the full report, and I'm looking forward to seeing the outputs from other countries too. But two inferences leap out for me, although the wording of the survey makes it hard to be 100% certain of respondent perception:It transpires that something similar is true of fibre in the fixed world - especially in FTTP (fibre to the premise) guise, according from some proper consumer research done by friends at Diffraction Analysis. People like raw speed, are annoyed when they can't get it, and are prepared to upgrade further once they do. They might even use it to consume extra operator-enabled services as well.
One of the regular Twitter debates I have with colleague Martin Geddes is whether broadband "performance" is about specific, measurable, application outcomes or not ("QoS" to most of us). He links that to an argument for non-neutrality, as he contends that specific Internet application providers have different performance requirements and should be allowed to "trade" for them, given the limits of the network.
Conversely, I argue that a lot of the benefit and value from fast, open (and neutral) networks is not application-specific, but rather is intangible. It's the potential for both users and developers to do what they want today or in the future, without trading or risking potential competitive abuse or extra friction added by ISPs.
It's like having a powerful car but sticking (mostly) to the speed limit - you know the "shove" is there if you need it. That extra acceleration on the highway on-ramp doesn't get you to the destination faster - but it feels good. Some even pay a premium for unreliable or twitchy supercars, despite the chance of a bad (even catastrophic) "outcome".
We're human. Perceived performance - and flexibility - is often more important than measured performance. Doesn't matter if it's broadband, cars, fashion, headline megapixels in a camera or a million other areas of life. There is almost always some correlation between "More of X" and "Better outcomes", but even if it's not a perfect correlation, we don't care.
Purists myopically focused on "optimisation" often don't understand this, or other human emotions like agency/control, a sense of novelty, image and so forth. While these apply more to individuals than businesses, there other factors like privacy and security and agility too.
In other words, "just give us faster networks & then get out of the way. Occasional glitches are a price well-worth paying for freedom and permissionless connectivity".
It's true of 4G - most people upgrading are happier than with 3G, although obviously some of that is down to the device itself, or perhaps better coverage with 7/800MHz vs. 2100MHz on some networks. Many notice the initial "whoosh factor" - that it's just fast even if they're doing something that doesn't need it. Sports-car feeling, again. And over time, developers are exploiting that, even if they know that there will sometimes be issues they have to work around
And that's the thing - the vagaries of mobile networks (coverage, congestion and so on) have taught application and content providers to work around performance limitations and occasional failures. To expect them and plan for them - so they use cacheing, variable bitrates, UI interventions to warn users of problems, clever codecs and error-correction. There will always be a weakest link. It's not perfect, but it beats having the hassle, friction and perhaps commercial conflicts-of-interest involved in paid QoS. It's why "non-neutral" mobile business models won't succeed, even if the law allows.
But back to fibre and FTTx, and some hard data. I don't often reference other analysts and consultants. (Martin G is one that I collaborate on about voice/comms, and disagree with on networks). But when it comes to the fixed-broadband world and especially fibre networks, access business models and wholesale metro, I'll gladly defer to Benoit Felten, who's been covering that beat for years. He now runs his own research firm, Diffraction Analysis, originally out of France although he's now living in China.
Many telcos have been slow at rolling out fibre, often because they have been unconvinced that consumers really want it, would pay more for it, or might adopt additional services as well.
Diffraction has been doing some interesting work in collaboration with the FTTH Council Europe, looking at the real-world experience of consumers who have fibre broadband, and comparing it to ADSL. Benoit has now published research on Sweden (one of the most developed FTTx markets in Europe), and has been working on France and Portugal studies as well.
I've had a chance to have a look through the Sweden report, and it corroborates my views in various ways (although it doesn't tackle neutrality, per-se):
- Fibre is perceived to be higher-quality than DSL, even by people who don't have it.
- FTTP users are more "satisfied". And higher speed FTTP equates to even more satisfaction. It's not about individual "outcomes" specifically.
- Fibre improves satisfaction with various "speed" metrics - latency, upload, download, variability and so on.
- Individual users are happier with FTTx when asked to compare with their past DSL experiences.
- Most DSL users perceive fibre to better (presumably because of friends who have it, or media coverage).
- Upgrades, both DSL-to-FTTP and FTTP-to-faster-FTTP are typically linked to wanting more "performance", ie speed.
- Lots of DSL users won't be upgrading soon, either because they can't get FTTP where they live, or because it's perceived as too expensive.
- FTTP correlates with higher use of triple- and quad-play, although it's not 100% clear which is cause & which is effect here.
- FTTP users do more stuff like streaming, video-calling, VoIP, tele-education etc.
- FTTP users seem interested in advanced services (perhaps with operator involvement) like telemedicine, digital home services, TV videoconferencing etc.
- People like fast Internet access, for its own sake. Speed sells, and feels good irrespective of specific applications or outcomes.
- People who like fast Internet also seem more interested in possible non-Internet network services too.
To me, this suggests that not only is there a business-case for investment in faster networks (FTTX, 5G etc) but that we need to consider both measured and perceived performance. Tangible and intangible. This is something missed by most of the economic-led studies on broadband - and certainly by all those debating the FCC's Title II Net Neutrality plans this week.
The Diffraction Analysis full report (32 pages) titled "FTTP Dynamics in a Mature Market - Swedish Quantitative Analysis" is available in two versions:
- 1-3 users (€1000 - click here)
- Corporate licence (€1500 - click here)
Contents pages are available on request via email:
information AT disruptive-analysis dot com
The links are to Diffraction Analysis' billing (although my company Disruptive Analysis has a financial interest here). You should get the report emailed through within 24hrs (NB the time difference given Benoit's location in China).
Note: if you're based in France you'll need to add VAT - if so, or if you want to pay by a method other than Paypal/credit card, or get more details about the report please get in touch via information AT disruptive-analysis dot com
(Note: I wouldn't be recommending research if it wasn't thorough, interesting, and in analytical coherence with my own view. However, it's Benoit/Diffraction's product, so the T's and C's are not my own)
No comments:
Post a Comment