October 15, 2009

Development in software business models – and the power of hindsight

When I recently was selecting a mail solution for my new company, I went for Google Apps. I did this after also considering a small business edition of Exchange, as well as just continuing to use the built in email service from my domain hosting provider that I already had up and running. Also my current client, i.e. my previous employer Inspera, has moved to Google Apps from an in-house Exchange solution, they just finished the transition last weekend.

These two decisions made me think back to what I wrote in my thesis on software business models in 2007 (part of my MTM degree, an MBA in technology management from NTNU, MIT Sloan and NHH - link to full foreword/abstract is provided for anyone interested):

…[even if] there is no “one size fits all” for business models in the software industry, if I had to place one bet on which business model innovation that will have most impact based on my findings, my decision would be clear: My money would be put on SaaS – on-demand subscription based software. Predicting paradigm shifts is a risky business, and especially so with shifts that have been predicted similarly without really happening before. Nevertheless, my take is that enough enabling factors have changed so even if it will not replace the traditional license, it may radically change the way software is sold, distributed and developed the coming years.

So the million dollar question is: Is this actually happening for mainstream applications for mainstream companies as we speak?

Extrapolating from the two samples (yep, two years since I last touched academia and reading the daily Dilbert since...) the conclusion has to be: “Sure”. On the other hand, I have also worked for some big companies over the last year that just wouldn't even start to consider Google Apps. But still, major changes in buying patterns often start with smaller companies and over time becomes an option also for the larger enterprises requiring time-tested and mature solutions.

Even if I feel the last two years have added support to my hypothesis of a possible “disruption from below” ala Christensen, I think the jury is still out on the question if SaaS will replace the traditional license based model as the dominant model at any point – or just be a supplement. (And to me financing a traditional license – making it look somewhat like a subscription – does not make it SaaS, by the way.) Cynics might also add that the predictions of net-PCs given in 1998 are not much different from the SaaS and “everything in the cloud” visions of today, they might argue that it has been coming in large scale “real soon now” for over ten years...

Reading the abstract two years after also drew my attention to another thing I wrote, that ...lines between data, applications, software and services can sometimes be blurry. This I at least think is even truer today than it was in 2007 – or to ask rhetorically: Is Twitter worth a zillion because of its software?


  1. Another piece of anecdotal evidence towards the added maturity of "the web as a platform" that I just thought of: When I wrote this piece I was using different browsers (uploading a sketch in Chrome on Windows, finishing it off in Opera on Linux) and did not even reflect that much over it. It was just on the web.

    On the other hand: At the "spark in eye"-stage, this article was a few thoughts jotted down using traditional locally installed software (on a netbook on a plane, IIRC - a place that is soon also to get widespread Internet access, it seems). Anyway, I guess personally I am very much in hybrid mode and will still be for a while...