What`s driving demand?

What’s driving demand?
A brief history of big ideas
End users don’t ask for much; all they want is everything, anywhere, instantly.
Demand has gone exponential. So how do we keep up?
To answer this, we first need to understand the big ideas that are driving
demand.
In the beginning, there was Napster.
When Napster was launched, the world was more concerned with the Y2K bug. It let
people share music for free and it grew quickly by word of mouth. In today’s terms,
the end-user experience wasn’t great. It was limited by the connection speeds of
both the uploader and downloader – and these were most often dial up connections.
But people didn’t care. At its peak, it had in excess of 50 million users. It was hailed
as the fastest growing business ever – unless of course, you define a business as
something that makes money.
The free business model is not much of a model if it’s your songs being given away.
Metallica and Dr Dre objected and the courts agreed. In 2001, Napster was shut
down and the music industry breathed a sigh of relief – that soon became a gasp.
The genie gets loose
Napster was gone, but it had spawned a new generation of peer-to-peer
technologies that couldn’t be shut down. Whereas Napster had been a single,
identifiable target, the music industry was now faced with a war on many fronts. It
launched 16,000 lawsuits, but this did little to stem the flow.
These new file-sharing technologies took advantage of faster home Internet speeds
to download songs faster. But faster downloads also started to increase end-user
expectations. If one file downloaded faster than another, they noticed the slow one.
People started to become impatient. They wanted their music faster, and as it turned
out, they were willing to pay for it.
Comparing Apple to peers
Even though the music industry was in a tailspin, the big labels weren’t willing to sit
down and make a deal. It took the force-of-personality of Steve Jobs to finally get
them to make an agreement that would enable legal music downloads.
In 2003 the iTunes store opened. It offered hundreds of thousands of songs from the
major labels. It wasn’t free, but it was legal and it was fast. Apple’s delivery
infrastructure effectively removed the uploader bottleneck from the equation.
Download speeds were now limited only by the downloader’s connection and songs
now took minutes to download instead of hours. iTunes sold 1 million songs in its first
week and has sold over 10 billion to date.
Bandwidth takes off
But as much as the success of iTunes was about Steve Jobs getting the labels
onboard, it was also about the increasing speeds of home broadband connections.
When Napster was launched, only 3% of US homes had what could be called
broadband connections. By the time iTunes was launched, this had reached 18%.
And every type of connection – broadband or dial up – was getting faster by the year.
Nielsen’s law states that the speed of a high-end user’s Internet connection will
increase by 50% each year. This has
held true for the past 30 years.
Of course this increase is not
universal. Ironically, Apple cofounder
Steve Wozniak said recently that he
still doesn’t have a broadband
connection to his California home.
Streaming hits the spot
By the mid 2000s, connection speeds
were fast enough that not only could people download songs, they could also stream
them live at listenable quality. Internet radio provided a real-time listening experience
but users were still limited to listening to what was being played.
In 2008 Spotify launched a service that took streaming to another level: instant music
on demand. It was like Napster, but instant. And like iTunes, it had the backing of the
big music labels. Today, Spotify offers over 16 million songs.
Faster Internet access had made instant musical gratification possible. With it came
a quantum leap in customer expectation. Now that they could have music instantly,
there was no going back. But music was only half the story.
We can't rewind, we've gone too far.
Having allegedly killed the radio star, video had moved onto the Internet.
Since the mid 1990s, it had been possible to stream videos. But on a dial up modem
it was more an exercise in patience than entertainment. In 2005 YouTube was
launched. It allowed people to not only stream videos, but also to upload their own. In
this way they created their own content and avoided Napster’s fate. Soon cat-lovers,
end-users, and companies all over the world were rushing to put their videos on
YouTube.
OMG I’m on YouTube!
In the early days, YouTube was as much about watching the buffering bar as it was
about watching the content. To improve the experience, YouTube adopted content
delivery network (CDN) technology to host their most popular videos in multiple
locations around the world. This enabled them to connect to customers via the most
direct path and the fewest hops. Just as the shortest physical distance between any
two points is a straight line, the shortest virtual distance and delivery time between
content and customer can be achieved by a direct connection.
However, less popular videos were hosted in fewer locations and often provided a
lower quality experience. End users naturally compared. If anything, as home
Internet speeds increased, end users were becoming less satisfied, not more.
But it was still free, and by the end of 2009, YouTube was getting one billion views
per day. One billion!
By 2013, YouTube accounted for 20% of all US Internet traffic during primetime.
But even that didn’t make it the biggest.
Buffering gets the flick
While it was obvious that the world would never get tired of cat videos, end-users
wanted more. They wanted the premium content that was on TV and at the cinema.
They wanted it in high definition, and of course, they wanted it instantly. The market
was ready for a premium model.
Netflix was used to doing things differently. In 2007, they sent out their one billionth
DVD-in-the-mail. Three years later, they had transformed into the largest source of
primetime Internet traffic in the US. By 2013, they accounted for 30% of all US
downstream Internet traffic during primetime.
In terms of numbers, Netflix streams fewer videos than YouTube, but they are
generally much larger files. Standard definition Netflix content streams at 3850kb/s
and HD content at 5800kb/s. And with ultra high definition (UHD) standards just
around the corner, this is only likely to grow. Multiply that by a lot of people, and
that’s a lot of traffic.
At any given time, the Netflix library
contains more than 3 petabytes of
movies, TV programs and video content.
Making sure this is all available on
demand, for customers all over the
world, requires some serious delivery.
Initially, Netflix used a variety of delivery
networks, but eventually economies of
scale meant that they were better off
building their own.
In 2012, Netflix launched the Open
Connect platform. This is offered to
Internet service providers (ISPs) for free
and encourages them to deploy Netflix
caching servers directly on their
networks. This lets Netflix preload their
most popular content into local caches
during off-peak hours, ready to be
downloaded. While this reduces the
need for Internet transit for ISPs, the load remains on their access networks.
Netflix relies on serious bandwidth – it would have been useless when Napster first
came out. The same is true of all of these big ideas; they only work because
connection speeds keep getting faster. But as demand continues to grow
exponentially, it takes something special to keep up.
It’s a bird. It’s a plane. No, it’s a...
Super-channel. That’s the word that is keeping network providers ahead of demand –
at least for now. We have come a long way from dial up connections on copper lines,
when success was measured in kilobits. The introduction of optical fiber meant that
by the late 1990s, wavelengths could be used to deliver up to 2.5Gb/s. In the decade
that followed, this grew to 10Gb/s, 40Gb/s and then to 100Gb/s. In 2012, Infinera
took the next step up. Over a section of fiber on TeliaSonera International Carriers’
backbone between San Jose and Los Angeles; they sent the world’s first terabit
transmission. It used 2 x 500Gb/s line cards and signaled the beginning of the super-
channel age. Within a year, this technology was being commercially deployed. In
2014, 1Tb/s transmissions from a single card look destined to become a reality.
Super-channel technology will let carriers stay ahead of bandwidth demand for now.
But while size matters, there is more to demand than that.
Bad ping kills the game.
With even the heaviest on-demand video, once the stream is in place, it’s one-way
traffic. Gaming on the other hand brings latency into the equation. Milliseconds mean
the difference between life and death online. And gamers aren’t known for their
patience.
When two players are sharing a console, latency is not a problem. But when millions
of players all over the world come together to do battle online – that’s a whole
different game. For low latency, a short path is crucial. And the best way to achieve
that is with a direct connection. Best effort networks will route traffic halfway round
the world if it gives them a cheaper routing. Carriers who own their own global
backbone don’t have to compromise. They can provide a direct connection between
players on either side of the world. This ensures the low latency and high Quality of
Experience (QoE) that gamers demand. And it’s not just gamers either. As more of
our everyday life moves into the cloud, low latency and high QoE are becoming an
everyday necessity.
The spoiled generation?
In not much more than a decade, end users have gone from being happy
downloading songs in hours, to demanding instant-response-HD-everything.
Because this has been a gradual progression, they haven’t really noticed. Wanting
more – and getting it – is all they know. We have created a spoiled generation. They
want everything – and we need to keep giving it to them.
We don’t know what the big ideas of tomorrow will be. But it doesn’t take much
imagination to see that they will require more bandwidth and greater performance. To
deliver this, carriers need to keep investing in technology,
To keep up with demand, carriers should be investing in technologies such as superchannels to deliver more for less. And for performance, they should keep expanding
their backbones, to make connection more direct.
We may not be able to predict the future, but we can be ready.