When launch success highlights a lack of forward planning

At the risk of showing my age, and giving away just how long ago Cactus Outdoor was founded (30 years this October, just FYI), let me tell you a story of technology in days gone by. Back in 1995 when Cactus first shifted from its birthplace in Wellington to its new home in Christchurch, we decided that having an email address would be handy. The fact that my then girlfriend and now wife was back up in Wellington while I was in the south may have played a part as well, but I digress.

Back in those days an email address was relatively rare. As for a corporate website, in our industry it was genuinely unheard of. Cactus founder Gwilym, in a move of inspiration, decided that it would be worth teaching ourselves to code  and to build our own website. Cue a bunch of experimentation and some old-school web hosting with early Christchurch Internet provider Planet.

I was harking back to the old days recently when I witnessed the launch difficulties that a particular internet-based company experienced. The sector that this organisation works within isn’t really relevant to this story. What is important, as we will soon see, is that it was in a space that was likely to see big spikes in traffic due to it being a fairly likely candidate for widespread media coverage.

The people behind the business were very smart and had created a solution that had very real applicability and a high likelihood for great take-up. Unfortunately they hadn’t factored this into their technology planning.

You see, back in the aforementioned early days of technology, setting up a website meant buying a server and trying to estimate what sort of traffic would be needed. It was a game of crystal ball gazing and trying to avoid having to spend more money than was necessary. Back in those days, internet capacity was a big ticket item.

Also back in those days cybersecurity wasn’t a thing. We didn’t need to worry about nefarious parties trying to launch denial-of-service attempts, phishing campaigns and other such dastardly plots on our websites.

Fast forward to the present and, thanks to the creative thinkers sitting inside companies like Amazon Web Services, Google and Microsoft, increasing capacity for a website or webservice is a simple thing that often doesn’t even need human intervention. Thanks to a myriad of different product and service organizations in the cybersecurity space, and a growing awareness of cybersecurity risks – we can mitigate against those risks.

Those services all fall under the category of cloud computing, a term that is very much become common parlance. I still chuckle remembering when buddy Raj Manji, now leader of The Opportunities Party, introduced me to a rather portly gentleman who was formerly the Member of Parliament for Ilam. Raf introduced me as a global cloud computing expert to which our elected representative asked me if I worked for the Meteorological Service – ah democracy, such a wondrous thing.

Anyway, back to computing. We now live in a world where infrastructure is, in tech parlance, available and controllable programmatically. In language we commoners can understand, what that means is that the days of old, where people had to stack physical servers one on top of each other and do technical stuff with wires and things, are over. Today all that is needed is a few simple lines of code, and internet capacity can miraculously ramp up and down according to demand.

It’s the reason that services like Google, Facebook and well-built internet banking solutions can handle huge spikes in load and still remain available. The corollary of course is that there is little or no excuse for a website to be unavailable – there are a bunch of tools available that can alert management to outages or issues (hat tip to awesome Wellington tech company, Raygun) and the aforementioned cloud computing vendors who can do all the heavy lifting to ensure automagically scalable infrastructure is available anytime.

In this case, however, the heroes of our story made two mistakes – firstly they didn’t plan for huge spikes in customer demand and hence computing load. Secondly, they failed to get really good advice from a trusted cybersecurity partner (tipping my hat in the direction of State Owned Enterprise Kordia) to ensure that the nasty guys couldn’t do bad stuff to bring their service down.

I’m not wanting to cast a shadow on the subject of my article – hence not naming them. But it is a cautionary tail for anyone who wants to deliver a message or a service via the internet – the great thing about the ‘net is that one can have virtually limitless reach. The bad side of the internet is that incredible reach can do crazy things to your demand curve. Fortunately with the advent of cloud computing, that’s an issue we need no longer lose sleep about.

Ben Kepes is a Canterbury-based entrepreneur and professional board member. He was a cloud computing evangelist before most knew what the cloud was.