Data is ‘the new oil’ in an increasingly digital economy, with the practice and power of data analytics being exalted – rightly so – as one of the most transformative levers a business can pull to gain insights, improve efficiencies, manage and reduce risks, and increase productivity.
By David Cosgrave, business operations director at SAS
Data science as a profession is enjoying it’s ‘moment in the sun’, with its practitioners in high demand.
Simultaneously, programmers and data scientists in the analytics industry (both globally and locally) are looking for – nay, expecting – increased flexibility and speed from their tools, so that there is never an opportunity missed to make that critical data-driven decision or change course when new opportunities arise throughout the analytics journey.
This demand is one of the driving factors in the explosive growth of open source software and programming languages in the analytics space. In short, because of its open development nature, it provides that elusive flexibility. That’s not to be sniffed at.
Not a quick fix
But, there is a rub: open source also comes with its challenges. One major drawback we often hear from data scientists and developers is that most open source is less user-friendly, and can require the intervention of specialists (dedicated staff, or ongoing training) to operate properly. At the same time, by its nature, open source usually lacks dedicated tech support for bug fixes and maintenance. Hiring in an expert can be costly, and time consuming; so when an open source solution falls over, the drawbacks somewhat negate the advantages.
An additional concern to consider is the auditability and replicability of results from certain open source solutions. This is and should be a major concern for any corporate, especially those operating in highly regulated industries like financial services and insurance.
No free lunch
It is these types of restrictions and limitations that underpin one of the most well-known sayings about open source – as free software movement activist Richard Matthew Stallman (aka ‘rms’) said, “think free as in free speech, not free beer”. That is ‘liberated’, as opposed to ‘no cost’.
Liberated programs and tools come with the power of the crowd. They fill common gaps and address widespread problems. Open source is most suited to non-differentiating components, while proprietary is best suited to innovation and custom functionalities.
Proprietary software comes with accountable support, often simplified UX, and ‘audit-ability’ cooked into the mix. In practice, companies could use open source software as a cheaper option which cuts down development time, and customise it to innovative proprietary software in order to match business-specific needs.
Best of both worlds
We believe the best solution is one that combines the good of both, and addresses the failings of either.
Although the two were previously very competing strategies, more and more we are now starting to see the value in getting “the best of both worlds”. As SAS, we are aware of the rise of open source and respect the flexibility that it allows. With this in mind, we are working to orchestrate and enable your analytics journey, through the provision of both choice and control – enabling the ability to gain the benefits of open source, while addressing its major drawbacks.
This is achieved through partnership, a system where businesses are not constrained to using Base SAS, and are able to choose their own programming languages – essentially using SAS in conjunction with whatever open source languages they’d like, be it Python, Spark or Java.
SAS firmly believes that the combination of traditional proprietary software and open source will not only help to create a sustainable analytics ecosystem, but one that supports more digital transformation goals – and we intend to be a supportive partner on this journey.