Original article ISPreview UK:Read More
A new report from analyst firm Enders Analysis, which was recently shared with ISPreview, has revealed that broadband traffic volume growth across most developed countries has “slowed to a relative crawl” to become the “new normal” (i.e. falling from 30%+ for many years and now at 10-15%). The story is similar for mobile (4G, 5G etc.) services (falling to 5-10%).
The study – ‘Low growth needs a new approach‘, which examined data/internet traffic across several developed markets, including the US, UK, Germany, France, Italy and Spain, noted how there had been many years of fairly strong and consistent growth until around 2020. Traffic then surged during the COVID-19 pandemic as people locked themselves away at home and worked remotely, but then sharply dropped back in 2021 to correct itself.
However, rather than return to the level it was at before, traffic volumes across both fixed broadband and mobile (mobile broadband) connectivity have instead continued to fall and remain significantly below pre-pandemic levels. This is more than just a post-pandemic correction, and is now said to be the “new normal“.
“Telcos (and governments/regulators) would therefore be advised to be prepared for this slower growth to be the new normal, and break away from previous habits of perpetually referring to explosive traffic growth and assuming that the latest high-capacity telecoms technologies need to be deployed everywhere for a high-tech economy to be able to function,” said Enders.
Broadband subscriber growth has also slowed to 0-3% post-pandemic, although this is perhaps less of a surprise given the huge impact that COVID-19 had on society (trigging lots of people to get a broadband service installed or upgraded) and the maturity of modern network coverage. For example, in the UK 30Mbps+ connections now cover 98% of premises and gigabit broadband reaches 86% (here), while there’s only a small portion of the population left to get online.
The issue of slowing traffic growth can also be partly linked back to the limited development and take-up of higher quality video streams. Most consumer internet traffic is generated by video content (IPTV, Streaming etc.) and so any developments in this field can have a big impact on traffic volumes.
On this front we haven’t seen much forward development since the introduction of 4K (Ultra HD), which is partly because higher quality standards like 8K and beyond are largely irrelevant to the screen sizes people are still using these days (Smartphones, laptops, tablets etc.). Streaming services often also continue to price 4K content at a premium, which suppresses take-up.
At the same time, online video providers have been adopting more efficient codecs, which compress those higher quality videos into ever smaller data packets that use less bandwidth to deliver the same stream – often significantly less.
Enders notes that other bandwidth-hungry services, which were once expected to help drive future growth in internet traffic, have failed to deliver. For example, cloud gaming has been about to take off for years, with many failed attempts, but most people still prefer to own their own copy of a game (digital or physical). Virtual reality, augmented reality and virtual presence have also been slow to take off, despite much hype a few years ago.
Enders Analysis Statement
Looking forward, there are a number of drivers of future traffic growth, which may well lead to spikes in individual years, but appear unlikely to lead to the sustained 30%+ per annum growth of the past. Firstly, live TV is (slowly) migrating to the internet in many countries, with platforms such as Sky Stream and Freely in the UK looking to make this transition seamless from a user experience perspective, and allow cost savings from (eventually) shutting down legacy broadcast platforms.
Live TV still has very significant volumes (around 40% of video viewing in the UK), and any sudden shifts would cause serious traffic spikes, but given a likely transition period of 10-15 years at least, this would not support sustained 30%+ per annum volume growth.
The impact of AI on traffic volumes is varied and still highly uncertain, with inter-data centre volumes already significantly impacted, and the impact from AI-based bots and AI-enabled video sensors could be very significant. However, online video traffic consists of hours a day of high bandwidth traffic to pretty much every individual in the country, and no other application (even AI-supported) looks likely to be able to match this traffic load, let alone exceed it to the extent required to drive high growth for many years.
The flip side of this, as touched on earlier in this article, is that the capital expenditure requirements on network operators will be “very much lower“, with “continuous capacity upgrades no longer required“. But we think it might have been more accurate to say that they’d take place more gradually, over a longer window of time, rather than simply being “no longer required“. The catch is that a lower frequency of network upgrades (i.e. core capacity rather than local access technologies) risk making any existing network deficiencies more obvious.
All of this seems unlikely to have much of a negative impact on the rising take-up of modern full fibre (FTTP) broadband networks, where many consumers are often just happy to get a reliable service that can actually deliver on the promised speeds (something older copper ADSL/FTTC networks often struggled to do). Not to mention the inevitable withdrawal of copper-based lines, which over the next few years will push everybody on to optical fibre.