Over the past 10 weeks, our experts have been in conversation with Internet of Things (IoT) deployment leads to understand critical success factors and challenges. One major finding stands out. Over and over again, we heard the same message: projects are at risk because of a shortage of data science skills.
Organisations simply cannot find the data management and analytics skills. They need to take advantage of their IoT deployments. Analytics is a key part of exploiting their new resource, and there is a worldwide shortage of data scientists. But it doesn’t have to be this way.
In our view, there are three ways that organisations can acquire the skills that they need to seize the moment: they can build, borrow or buy.
This is perhaps the most obvious: customers can develop their own staff, to ensure that they have the skills that the company needs. But while most of those interviewed could articulate the skills that they needed, very few were able to suggest how they could support their staff to develop those skills. Data science is not something that is widely taught as yet.
It is precisely to address this need that back in April this year, we launched the Academy for Data Science. It provides certified training for data scientists. Our customers and others can use this academy to create an in-house development programme for their staff, to provide skills tailored to the company’s need. The six-week training modules include theory plus case studies or team projects, coaching and an exam to achieve certification. Although the course focuses on SAS packages, trainees emerge with a general certificate in data science.
The second option is to borrow the skills from elsewhere. We found that a number of companies surveyed had gone down this route, forming partnerships with firms that had the necessary skills in developing and handling IoT technology. Professional services firms like SAS have access to experts in analytics and data handling. What’s more, these firms ensure that their experts’ knowledge is kept up-to-date by regular training and professional development. It is a time-effective way of ensuring access to the latest skills and technology.
Of course the big drawback of ‘borrowing’ skills is that eventually you have to pay back the loan, as it were. The borrowed experts have to leave and move on to somewhere else. But a combination of borrowing and building can pay off in the longer term. Borrowing fills the immediate skills gap, and building then starts to take over. What’s more, ‘home-grown’ staff developing their skills through training and certification programmes can work alongside borrowed consultants for a while, shadowing them and learn ‘on-the-job’ as well as during their training sessions.
Buying into capabilities
The final option is to buy. This could be in the form of recruiting skilled data scientists who will be ready immediately. The problem with that, though, is that they are few of them. There are not enough to go round, and anyone wanting to recruit has to consider what they have to offer to attract and retain this rare breed.
Fortunately, there are other options: to buy pre-configured solutions that reduce the need for manual intervention by skilled professionals. There are several ways to do this:
- a plug and play module
- a more dynamic data lab
- an online ‘cloud’ analytics-as-a-service capability.
For example, the Analytics Fast Track™ for SAS® (AFT) is a‘plug-and-play’ analytics module designed to enable businesses to get value out of it fast. The idea is that business simply turn on, add data, and start to benefit immediately. Built in partnership with Intel, this suitcase styled configuration comes pre-configured and ready to go.
For teams who need to support many and as yet unknown projects, a big data lab might be a better option. This is an environment designed for experimentation and ‘fast failure’, and to generate value from a very early stage. It comes with requisite data science support elements so teams are not stuck at any point.
The ultimate in the ‘buy’ spectrum is a fully functional cloud capability. SAS Viya™ for example provides high-performance cloud-based visualization package designed to be accessible and scalable for individual business users. It is suitable for any analytic challenge, large or small, and also helps with technology integration.
The hybrid model will be the most sustainable for Data Science
Immediate skills shortages do not need to hold anyone back from exploiting IoT opportunities. Whether you decide to build, buy or borrow, or some combination, there are options out there. What will you choose? We can help you in that: join us on Friday 5th August for a discussion on analytics skills evolution. This is an open discussion on Twitter, and no registration is requires. Just follow the #saschat hashtag which will be most active between 15 and 16 hrs CEST.