Like anything in this world, data science has two sides of the coin, and you cannot have the one without the other.
On the one side, we have the bright future data science is promising us, enabling organisations to gain valuable insights from data. This helps to innovate and build great systems to improve business and revenue with the ultimate goal to have all processes and analytics automated with AI, Machine and Deep Learning. All this by capturing and using real-time data for real-time analytics. This looks great and organisations are jumping to get to this as soon as possible.
Before getting to the good side, organisations must deal with, and work through, the flip side which currently consists of quite a few challenges and issues. Luckily these challenges are shared across multiple industries and sectors, which ensures they are being addressed and communities of data professionals are getting together to share knowledge and insights to overcome obstacles
When I spent time with 20+ leading data analytics professionals at the IBM & Corinium Data Science CoE breakfast in September, I found that major challenges range from culture, building and retaining the team, data ownership, overflow of vendors, data privacy and security, translating/visualisation, budgets, unclear or no strategy to a disconnect between data / IT / business teams as well as internal politics. Just to mention a few
Data Scientists are struggling to do what they do best, that is delivering true data science. If Data Scientists have the support they need, such as frameworks and tools in place to take care of governance, they can innovate and deliver as they should. More pressure is being thrust uopn the scientists to deliver faster, and governance is being pushed down, which leads to decreased data quality and accuracy.
How can the speed of delivery be increased?
One enabler is the cloud. Organisations are starting to investigate this, but one major concern when having your data in the cloud and doing analytics in the cloud, is data privacy and security.
Implementing your ‘sexy’ strategies and projects can be difficult and frustrating if you don’t get the buy-in and trust of the organisation. Data Scientists can come up with the most amazing solutions which could benefit the organisation and take it to the next level, but if they cannot communicate this to the finance team who allocates expenditure, it is not worth anything. Data visualisation experts are of utmost importance to communicate and translate the value and benefits of these data projects to the business professionals. This is to get their buy-in, support and budget allocation to test and implement these projects.
Data Science is indeed as ‘sexy’ as we think, however, it is a new practice and has a long road ahead, but organisations are jumping on this and moving forward. Some are moving forward cautiously, and some believe moving fast, failing and adapting quickly is key. Still, you can only move as fast as your level of data maturity, resources and current skill sets allow.
Culture eats strategy for breakfast, so look at getting the culture and team in place first, the rest should follow organically and prove to be more sustainable.
Gert Botes is the Portfolio Director: Data Analytics - MEA and is in the process of producing DataCon: Africa's Most Trusted Data Analytics Conference. Find out more here.
