I have been in the business of delivering software applications and solutions that have advanced and predictive analytics at their core, in one form or another, for last 25 years. In a recent conversation with colleagues we were discussing how the wider market has changed and, more specifically, how this has affected the way that people who analyse data, create predictive models and help their organisations use them go about developing their skills and keeping them up to date.
It feels to us that the dramatic developments in the availability and use of technology have had an equivalent impact on the skills development market (what we used to call training!). In this blog I’ll discuss the factors we see driving change for those using advanced and predictive analytical tools. I think there are five key trends.
The explosion of open source technology
The last 10 years or so has seen the emergence and relentless expansion of open source and ‘free ware’ options for advanced analytical tools. Languages such as Python and R have had a radical democratising effect on the analytical software market because everyone has access to them. They have also had a wider impact on provision of high quality, accredited training.
Open source programming languages designed for analytics are complex and broad in their capability. The downside is that when an aspiring user is able to download and attempt to use such tools with no immediate investment, it can be daunting to then have to ask their organisation (or even worse, put their hand in their own pocket) for funding of training.
The evolution of the self-taught data scientist
Following on from the point above, the emergence of open source analytical technology has led to the evolution of a generation of self-taught data scientists, used to developing their skills and solving analytical challenges by leveraging access to online communities, social media and forums, being able to borrow a snippet of code from here or there and then tweaking it to meet their specific requirements.
There is a lot to be said for this approach to learning and using new and powerful technologies. There are downsides though. It can lead to sub-optimal productivity for a user who can end up spending a lot of time searching for the answers to their questions, and may also result in analytical and modelling processes that are not the most efficient for a given problem or scenario.
A loosening of the grip of large corporates on analytical technology
Many very large software corporations grew large by being able to pretty much set their price for advanced analytical software and the associated professional and technical services, including training. The advent of the two phenomena described above has led to something of a backlash amongst the user community and a rush to embrace what is now a much healthier and more fragmented market for analytical tools and platforms.
Overall this is a much, much healthier environment and allows quality competition to flourish. The only downside I can see is the impact it has had on the provision of really good quality user training. The combination of open source technology, open resources that allow a user to self teach to some extent and the long history of over priced ‘corporate only’ provision has diminished the perceived value of investment in user training.
I firmly believe that lack of investment in training costs in terms of speed to good results and increases the risks (in an organisation) of early disillusionment with the use of advanced and predictive analytics.
The rise of cloud computing
The changing way we now access our technology, increasingly via cloud based platforms and evermore on an ‘on demand’ basis, has also altered behaviour. Organisations often feel reluctant to invest in associated training and enablement when learning how to work with, use and deploy predictive analytical tools and their outputs.
The impact is that we expect only to have to pay for exactly what we need, exactly when we need it. Not unreasonable when you consider the background and history we’ve been exploring in this blog post. For providers of technology and training services, especially those that ask users to pay, that has meant a fundamental and significant rethink of business and cash flow models.
The relentless pace of technological change
There is no escape from the fact that technology pervades almost every aspect of our lives today. Predictive analytics is a discipline and practice that is only possible through the use of the latest computing and technology. The explosion of tools and platforms, and the changing ways that we access and pay for them have been subject to the white heat of persistent change. This constant change makes the decision making related to investment of time and effort into technology based skills training even trickier. No wonder individuals and organisations are thinking carefully before investing.
In summary, it seems that the technology training industry in general and the predictive analytics training activities in particular have been hugely impacted by the wider technology shifts. I recall working in large US software organisations in the 1990s and 2000s that had large thriving and profitable public training businesses. I see little evidence of this model in today’s market.
Instead what we see is access to lower cost, bite sized and on demand training resources and extensive use of online user communities, social networks and forums. In short, a much more flexible, adaptable and faster evolving ecosystem.
As long as the quality is maintained perhaps we shouldn’t worry… on the other hand, I also see signs of skills shortages in the industry and I worry that perhaps there are still significant gaps in the resources available for developing the skills required. We have responded to this ourselves by developing a range of different training options to meet different user needs, from our range of free ‘how to’ videos, to ‘teach yourself’ user guides to bite sized online support and training, to tailored onsite group training.
