Modelling Internode’s Growth Rate

Internode is an Australian Internet Service Provider (ISP) based in my home town of Adelaide.  They are my also my particular ISP of choice having provided me with a fast, stable, reliable internet connection at a reasonable price over the last 5 years or so.  If you’re in Australia, and you’re looking for an ISP, I can certainly recommend Internode.

As a privately owned company, Internode is under no obligation to release any operational details such as number of subscribers, growth rates, financial statements and so on.  But every now and then they may choose to celebrate an achievement with a public announcement.  From these media bites at least some data can be gleaned.  Not enough for any serious analysis of course, but enough for a bit of fun.  I thought it might be interesting topic for my second blog post to string these bits of data together and see what model might be applied.

By way of background, as recently as April 2007 everything was booming along for Internode.  They had announced their 100,000th broadband customer, and were projecting this figure to double before the end of 2008.

But what a difference a year (and a bit) makes.  By August 2008 the company’s growth surge had, in fact, shown signs of abating.

With the booming broadband market expected to plateau during the next 18 months, Internode is preparing to change gears…

Internode is the largest independently-owned broadband services company in Australia. The company … now has more than 300 employees and 140,000 broadband customers nationally
Internode, August 2008

Which is still a great result of course.  140,000 customers isn’t anything to sniff at.  And it’s not like they’re going backwards.  But from the announcement it’s clear that Internode’s growth rate has begun to slow.  I’m not convinced that the entire broadband market is expected to “plateau”, but by their own admission Internode are now less sanguine about future growth.  At least in terms of increased subscriber numbers.  Internode’s strategy (or gear shift) now appears to be to grow revenue on the back of a static customer base by offering additional services.  That’s probably Business 101, but I was asleep in that class.

Anyway, below is a summary of the small amount of data on Internode’s customer numbers that I’ve been able to source from various places around the Internet.

date no. subscribers
4th quarter 2005 50,000
March 2007 100,000
August 2008 140,000

Internode, April 2007
Internode, August 2008

I must warn you that this is where my post veers off into the realm of gonzo statistics.  I’m going to use just the three points of data above and propose a model for use as a forecasting tool.  A bit dodgy, but as the famous statistician George Box said, “all models are wrong, some are useful”.  So with that in mind, let’s plug the numbers into the very cool Online Regression Tools website and see what it throws back.

Online Regression Tools makes the job ridiculously easy.  Converting dates to decimals, and ignoring the ‘000s, the following dataset can be copied and pasted into the Nonlinear Regression input box:

5.92, 50
7.25, 100
8.67, 140

Setting the “number of parameters” option to “2” Online Regression Tools recommends the following as the Best Function:

subscribers = 463.047 * loge(0.626 * loge[year])

Which looks a bit like this:

Good enough for me.  Solving for [year] the model suggests that Internode won’t see their 200,000th customer, originally expected before the end of 2008, until mid-August 2011.  But will we see a media release from Internode in early 2009 celebrating their 150,000th customer?  Time will tell.


One Response

  1. I think you have to look at the big picture, its more a population curve with a limiting function of lack of drinking water, caps on immigration and work for the dole scheme.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: