For the last week or so Chorus has issued a daily update on the data traffic passing through its network. After an initial surge when large numbers of people began working from home, things have settled into a pattern.
It’s the new normal. As Spark’s technology director Mark Beder points out, weekday data use now looks the way weekend’s looked before the lock down. Weekend peaks are now higher again.
Spark says the amount of data on its network has doubled since widespread remote working started. Data peaks are about 27 percent higher than before the pandemic arrived.
Peak mobile traffic is up 22 percent. The company says it has seen some congestion at times and is working on adding capacity.
Call of Duty update
An update to the Call of Duty game on the first weekend of the lockdown period caused what Spark technology director Mark Beder describes as a “massive spike”.
Away from Spark, Tuesday evening saw traffic peak at 2.70Tbps on the Chorus network. There was an update to the Fortnite game during the evening which may have accounted for the extra traffic. The busiest midday this week was Wednesday with 1.72Tbps.
Both Ultrafast Broadband and Enable Networks have registered similar increases. UFF says it now sees about double the amount of pre-lockdown data.
Traffic well within capacity, for now
The Chorus figures are well within the network’s capacity limits. In the run-up to last year’s Rugby World Cup, which was streamed by Spark Sport, Chorus and most of the rest of the broadband industry brought forward capacity upgrades by 18 months or so.
Today the network is built to cope with 3.5Tbps. That’s comfortably above the peaks we are seeing at the moment. Traffic could go higher again when school and university terms restart, but there appears to be more than enough headroom to cope.
New Zealand voice networks recorded the highest call volumes in history this week. There was a peak on Tuesday. This followed Prime Minister Jacinda Arden’s announcement she was raising the Covid-19 alert level from two to four and the country would begin a four-week lockdown.
All carriers experienced congestion. Callers overloaded government call centres with enquiries. The congestion affected the three mobile networks and Spark’s landline calling network.
A Spark spokesperson says: “Following the prime minister’s Covid-19 announcement today, telecommunications providers experienced call volumes beyond any level of calling ever seen in New Zealand. This is creating significant congestion for voice calling at an industry-wide level.”
The TCF spoke for the industry saying telecommunications companies were working quickly and collaboratively to fix the issues arising from congestion.
Carriers called on phone users to switch to digital communications technologies using the nation’s fibre network to free up voice lines.
Customer services struggling
Customer service teams struggled to cope. In part this is because operators closed their overseas call centres or are running them at a reduced level to protect staff from infection risks. There were also much high call volumes.
Vodafone issued a plea to its customers to use the company’s mobile app, chatbot and website where possible to reduce the load on call centre staff. The company asked customers were to contact the company’s social media team which extended their hour to cope with the extra demand.
The company says: “Due to precautionary measures in New Zealand and internationally, our customer care teams are managing the impacts of Covid-19 while dealing with higher call volumes. We have major call centres in different offices in New Zealand and India, and a small specialist customer care team in the Philippines – and while we’re able to redirect work and calls for some customers between them, we are also planning for future impacts including what we can expect will be further increased restrictions on movement in cities worldwide.”
Spark and Vodafone sell fixed wireless broadband as an alternative to fibre. The service is far from rubbish. Yet it isn’t up to the demands New Zealanders will make of it while spending more time at home during the Covid–19 outbreak.
Companies are asking employees to work from home. Schools haven’t yet sent students home, but may. Because of social distancing people stay at home instead of going to the movies, the pub or other activities.
All this means people will be more reliant on broadband connections. For work-from-home employees, broadband is their livelihood. For everyone else it is a useful communications tool. It also delivers entertainment needed to stave off the boredom of stopping at home.
All those examples need data. A lot of it. That’s one thing fixed wireless broadband can’t do. Most fixed wireless broadband plans come with data caps that limit a customer’s data. Which leaves people with a problem if broadband is the main Source of entertainment during the outbreak.
There are two other issues that suggest fixed wireless broadband is not up to the demands people will make of it.
First, fixed wireless is prone to congestion. A fixed wireless tower shares bandwidth between users. If there’s only one user at a given moment, connection speeds can be good. If everyone is online at once, the performance drops. It can drop below the level needed to sustain a video stream.
During the Covid–19 pandemic most people will spend the evening at home most of the time. The demand may swamp fixed wireless towers.
With everyone at home, the data traffic is set to beat the highest Rugby World Cup levels.
The second part of this is that with congestion, some users may not be able to connect at times. It’s one thing to miss the last 15 minutes of a movie. It’s another thing to have your main channel to the rest of the world shut off during the Covid–19 pandemic.
Which brings me to the third problem. Clogged fixed wireless broadband networks can impact mobile wireless reception and coverage. The mobile network is our vital lifeline. Few people still have copper voice line connections. If busy fixed wireless broadband towers make it hard to make phone calls we are all in trouble. We will need phones more than ever during the outbreak.
What should the telcos do?
First, stop selling fixed wireless broadband where fibre is an option. Get customers onto fibre where possible. Where there’s a mix of fibre and non-fibre premises, leave fixed wireless for people who can’t yet get fibre.
Where fibre is not available, it makes sense to give towers a capacity upgrade. Move to 4.5G or a higher technology. And then do a better job of managing the capacity. Which, means not milking the network too hard.
Fixed wireless has a role to play in the broadband mix. Pretending it competes with fibre diminishes its overall value.
Last week Spark announced its first half results for the six months to December 31. It is a solid report showing strong revenue growth.
Spark looks to be heading on the right track. Yet there is an interesting angle on one of the company’s strategic moves.
Nine paragraphs into the market release there is this quote from CEO Jolie Hodson:
“We made a deliberate decision to limit wireless broadband sales in the lead up to the Rugby World Cup, as a conservative measure to ensure customers had a great viewing experience while we introduced our new streaming service. Our capacity was more than sufficient, so we expect this to be a one-off and connection growth to return to trend in the second half.”
In other words Spark back-pedalled on fixed wireless broadband sales because senior management didn’t want customers to have a disappointing Rugby World Cup streaming experience.
Fixed wireless alternative
Spark pushes fixed wireless broadband to its customers as an alternative to fibre. It’s a strategic move because Spark owns its wireless network. That means the company doesn’t pay a wholesale fee to a fibre company. It keeps all the money and that makes for a higher profit margin.
Investors love that.
Downplaying fixed wireless broadband in the run up to the Rugby World Cup made sense. Although fixed wireless broadband should be able to give customers enough bandwidth to watch high definition streaming video, that’s not always the case in practice.
Unlike fibre, which has consistent and predictable performance, fixed wireless broadband performance varies from place to place. In some cases it also varies at different times of the day.
Fixed wireless broadband bandwidth is shared. So if a lot of people connect at once, speeds can drop. The Rugby World Cup saw data traffic peak across the nation. That put pressure on more marginal fixed wireless broadband connections.
Good at times
Fixed wireless broadband can be good. I’ve heard from happy fixed wireless customers who enjoy decent speeds and uninterrupted connections.
There are others who say their service does not do an adequate job with streaming video.
One common complaint is that wireless broadband speeds are not consistent. In some cases speeds vary in a regular pattern over the course of a day. Others say they get intermittent slow downs.
Conservative on fixed wireless broadband
Spark describes the decision to back-pedal on selling fixed wireless as conservative. That may be the case. But it underlines that the company is not confident about its fixed wireless performance.
There was no conservatism about selling fibre broadband to customers in the run up to the Rugby World Cup.
The message is clear: Spark knows fixed wireless broadband is a lower quality product. It knows customers get a better experience on fibre.
Customers with a 100 mbps fibre plan saw average download speeds of 99mbps. During peak time the dial barely moved. Samknows reported peak speeds at 98.6 mbps.
With fixed wireless broadband the average speed is 25.8 mbps. At peak times this drops to 22.7 mbps. That’s not a huge drop, but it squares with the anecdotal evidence that some customers see big drops while others see little or no drop.
Fixed wireless broadband latency
The SamKnows data also looks at latency. This is the time it takes for data to do a round trip. If latency is high, online users of applications like video conferencing and gaming can expect stuttering and dropouts. SamKnows says 30 ms is high.
SamKnows found nine in ten fibre connections had latency below 20 ms. In comparison 95 precent of fixed wireless connections had latency of over 30 ms. The average latency is around 50 ms.
Of all the latency tests performed on Fibre connections, 92% were below 20ms. At the other end of the chart, 95% of Fixed Wireless latency results were above 30ms.
That’s past the point where dropouts start. With everyday TV streaming, buffering can shoulder some of that load. Even so, it is a worse customer experience.
SamKnows’ summary says:
“…many fixed wireless connections will experience issues with latency-sensitive applications such as video calls and gaming.”
VDSL2+ can deliver near fibre speeds and in some cases is consistent and reliable. Before fibre came down my road I had a Spark VDSL2+ connection that delivered a consistent speed of more than 70mbps.
In three years it never wavered. You can read about my fixed wireless experience in this post. The speed was never anything like as fast as the VDSL2+ connection.
Fibre most reliable
Of course VDSL2+ is not as good as fibre. In the report summary SamKnows says:
“Households with multiple user should consider fibre, if available, for the most reliable performance.”
Spark knows all of this. The reason it pushes fixed wireless broadband is that the margins are higher. That’s because there is no wholesale charge.
For many Spark customers fixed wireless broadband is the right product. But let’s not pretend it isn’t an inferior product to fibre. Spark is willing to let its investors know that.
Disclaimer: I edit The Download magazine for Chorus as a contractor. It covers the company, the telecommunications industry and fibre broadband. These are my views and not those of Chorus.
Telcowatch says Vodafone is New Zealand’s mobile market leader.
There’s not much in it. Vodafone is one percent ahead of Spark on 36 percent.
The two were neck and neck for most of last year.
While the lead is real, it’s not dramatic.
Nor is it the whole picture. The way Telcowatch measures the market means that Spark’s Skinny business is counted separately from its parent company.
Adding that back into Spark’s figure puts the company well ahead of Vodafone with a 41 percent market share.
However you crunch the numbers both Spark and Vodafone have a clear lead on 2degrees. The third mobile carrier’s market share is stable at 23 percent. That makes it a little over half the size of Vodafone and Spark.
It probably suits everyone concerned to count Skinny as a seperate business.
Yet Skinny is definitely a Spark brand.
When Skinny started it was more distinct from its parent than it now is.
Today Skinny’s product alignment can be seen as rounding out Spark’s offerings. It’s a no-frills version. In supermarket terms it is PaknSave to Spark’s New World.
The two share the same network infrastructure. Skinny employees may be loyal to the brand, but they are Spark employees. Spark’s management decides Skinny’s strategy.
Skinny remains the smallest of the four brands. In December its market share was 5.6 percent. It has been between roughly five and six percent for the last couple of years.
The most interesting aspect of the recent report from Telcowatch is not the interplay between Spark and Vodafone, but the way Skinny has been growing its market share at the expense of the parent company.
Over the last year Skinny is the best performer in terms of market share growth. It has grown gradually.
It’s not hard to understand why. Despite all the fuss about 5G, the mobile phone market is mature. There’s less differentiation between brands and less of a premium in Spark’s brand when compared to Skinny.
There is, however, a considerable price difference. Slowly, but surely, customers are waking up to this. You can buy what amounts to the same mobile experience for less money. The big surprise is that more people have yet to realise this.