As analytics rose sharply to the top of the corporate priority list around the world, a new and very expensive problem emerged: there simply aren’t enough skilled analysts to meet the demand.
Three obvious solutions presented themselves.
The first is to train more analysts. But increasing the pool of core skills takes a long time — years, in fact. The industry and the higher education sector around the world have responded, but the benefits are yet to flow substantially. The other problem with this approach is that it misses a key pain-point: in fields such as finance or retail it’s often subject matter expertise — rather than technical expertise — that is hardest to locate.
The second solution was to outsource — often overseas — and this certainly overcame short-term skills gaps. When outsourced to very well-resourced markets like India. this also overcame the subject matter gap. However, it did not solve another problem: building analytics expertise into the brain-core of the business to reflect the reality that data analytics is now essential to how companies work.
The third solution — and the one the industry does best — was to take a big problem and turn it into software.
Ultimately this is where the market moved fastest, with the emergence of a generation of analytics software tools that put power into the hands of end user business analysts. These tools did some of the heavy lifting that used to require the direct hands-on engagement of scarce, highly skilled, and expensive data scientists.
A disruptable moment
In turn, that has put pressure on incumbent analytics software providers whose solutions were architectured for an earlier age. Those incumbents, it must be acknowledged, are awake to the shift.
At the top of this list sits SAS — for decades the default statistical and analytics solution and the must-have skill-set for a generation of data analysts. Which-50 asked SAS how it is responding to the circumstances in which it finds itself.
The company began by acknowledging the market shift, although it argues the commoditisation is primarily happening at the tools end of the analytics market.
According to a company spokesperson, “Tools, in this case, mean things like development environments and the code, algorithms, and methods used to build assets. This is being driven less by software companies and more by users of free, open source, particularly Python and R. Anyone can download Jupyter notebook and R Studio and free open source machine-learning packages from places like Github, and start coding and building models in minutes with very little effort.”
To adapt, SAS is focusing on ways to help organisations operationalise their experiments that use these open source tools, It does this in a framework it calls SAS Viya, which is designed to ease the process of publishing, scaling, managing, and governing those experiments in production — no matter what size the data and how real-time the processing requirements are.
“In addition, SAS continues to leverage its extensive industry and domain knowledge to deliver industry-specific solutions based on analytics for financial services, government, healthcare, retail, etc. as well as domain-specific solutions around customer intelligence, risk, fraud, cybersecurity, and the Internet of Things,” the spokesperson said.
The use of products like Tableau, Qlik, and Birst has also lead to a decentralisation of the approach to analytics — especially in the enterprise space — which in turns creates the problem of managing that diversity. The spokesperson told us, “Large organisations frequently say that they have two of everything, even the ones who have ‘standardised’ on a particular version! As of right now, there are few organisations that seem to truly have one enterprise standard BI tool. Often what we see are tools that have been adopted and standardised within a single department, like the risk group, or within finance.”
For the IT department, one approach is to support multiple tools, while trying to consolidate deployments and agreements to limit sprawl and hardware costs as much as possible. The SAS spokesperson says that company’s play “is shifting away from feature bakeoffs with BI vendors and tools, and focusing instead on infusing analytics and artificial intelligence into its visual interfaces on top of SAS Viya to help reach and empower the business analyst with analytics.”