Do we need shop software?
28th February 2 Comments
For a while now I have been thinking a lot about the general theme of possessing things, which already manifested itself in my recent musings on ownership. That article dealt with the difference between access and ownership and how this motif can be found in new business models. Here, I would like to try and apply this question to software, or shop software, to be more specific.
Those of you who have been following what I have been doing over the last couple of years know that I was and still am very much involved when it comes to onlineshop software. In the days of yore, it started with osCommerce and xt:Commerce, I delved into Magento – wrote two books on the software – and then had a closer look at OXID eShop which ended up in yet another book. Apart from those platforms, there is a whole bunch of other OpenSource and commercial software which interests me and which keep my business live, well, busy.
Software stack complexity
Usually, for all of these software packages, it works in a similar way: a core team works on the software, gets rid of bugs and adds new features. At more or less regular intervals, a new version is published which is then downloaded by the shop owner (or his agency respectively) and installed on the respective server as a fresh install or a software update. Take Magento as an example: According to the official blog, there have been more than 4 million downloads of the OpenSource version of their software so far. I’m not sure how this number gets calculated exactly, what counts as a download or not and how many of these downloads are eventually turned into productive online shops. The second number I would like to throw into the discussion comes from Builtwith.com, according to whom there are now about 75,000 websites in the world running Magento. (Again, those numbers have to be taken with a pinch of salt and we cannot be certain to get a 100% accurate picture.) For the mathematics geeks: the latter is less than 2% of the total downloads number.
Now let’s dive into this a little deeper. If we look at those 75,000 sites, this does not mean that all those shops are running on the same Magento version, oh no! From my own experience I can say that the software version can be anything between one of the first and one of the more recent. And Magento is hardly four years old – think about software that has been existing for decades! The older a software stack grows, the more heterogenous the picture will get.
Want some more complexity? For Magento, there are now about 5,o00 third party extensions which can be installed via the Magento Connect marketplace. And guess what, some of them have been around for a while, too, and exist in many versions. You can find any odd combination of different modules of different versions in today’s Magento installations. And to top off the complexity rampage, think about the underlying platforms the Magento installations are running on: Which OS is your server running on, which versions of Apache, PHP and MySQL are you using and how are those configured? Get what I mean?
The burden of ownership
The main reason for this is simply learned behaviour: since the mid-eighties, when Microsoft began selling its Windows OS, people were used to getting software in the form of physical data carries, be it floppy discs, CDs or DVDs. Since then, a piece of software usually comes in a box, containing one or more data carriers, a handbook and all sorts of additional swag. You pay the retailer, the retailer pays the software producer who in turn pays his developers. End of story. Once in a while, the end consumer would download a service pack and every two years or so buy another major upgrade. Thus, at the core of this model is the notion that users would actually buy the software and own something physical. On the producer’s side, this makes monetisation easy, because in regular, plannable intervals a new round of cash could be collected with a new update.
Let’s get back to the topic of shop software and the question, why this traditional model of software distribution as outlined above will work less and less. First of all, with modern broadband connections, those software packages can be downloaded fast because their zipped/packed forms very rarely exceed 100MB. More importantly, however, are diminishing development cycles/intervals. In times where everybody hails the real-time web, waiting an extended period of time for an software update becomes obsolete. Especially in the area of ecommerce, where one tries to outsmart the competition by using new-fangled features of brand-new software, speed is all that matters. A company such as Magento, OXID eSales or Shopware tries to publish a few major releases each year – just think about how many DVDs had to be sent around!
The majority of shopowners carry the burden of constantly having to update their software in a way that it doesn’t affect the rest of their IT infrastructure. Owning the software and having it run on one’s own server will increasingly become a luxury that, at best, can only be justified by the need of extreme individualisation. The latter, however, applies to only a fraction of active online shops.
New models for software access
For a while now, on-demand platforms and cloud-based services such as Demandware, Shopify, Rakuten etc. have offered their services and keep increasing the flexibility of their software to support an increasing number of business models and shopping concepts. Shopowners benefit from a stable platform without having to worry about exponential complexitiy of their own shop software installation. (And this is only one strong argument: in one of my next posts I will talk about why ecommerce offerings based on isolated software installations are disproportionally harder to market.)
But not only shopowners should learn a lesson, also onlineshop software producers will increasingly find their traditional business models under attack. When owning a piece of software becomes more and more unattractive, what does your portfolio look like in a few years from now?