Big data and analytics is set to create millions of jobs globally over the course of the next decade. One survey released last year by the Centre for Economics and Business Research found that growth in data analytics and the IoT sectors will create roughly 182,000 new jobs by 2020 in the UK alone, a number that will be dwarfed by US and China figures.
These new jobs do not, as some might believe, start and end at ‘data scientist’. Harvard Business Review notoriously labeled the role ‘the sexiest job of the 21st century’, and this has certainly borne out. Both their salaries and demand for their skills remain high. The Robert Walters Global Salary Survey for 2017 gave an estimated range for data scientist salaries of $116,000 to $163,500 in 2017, up 6.4% on 2016 levels. Data from recruitment website Hired, meanwhile, revealed that interview requests for data scientists has risen 33% since the second quarter of 2016 - the biggest jump for any tech role. However, while data scientists will remain in demand, the degree to which data has impacted every facet of organizations means that there are new roles being created all the time with the aim of ensuring it is being utilized to its fullest potential. Having a data scientist alone is not enough to guarantee a successful analytics transformation, and we’ve looked at some other hires you might want to consider.
There has long been a disconnect between data scientists and business users. Data scientists largely lack the business expertise to convert the data into actionable insights, while business users lack the technical understanding to do the same at the other end. Looking to bridge this gap are ‘data translators’.
As the job title suggests, data translators are there to make sense of the data for business users. They must speak both the language of the business and the language of data science in order to interpret the data for the layman and ask the right questions of the data science team. They must have deep organizational and industry knowledge, understand economics and finance, and be able to validate, integrate, and use advanced models within a broader decision support framework. Their background will probably involve mathematics or statistics, as well as functional expertise gained from experience specific to their industry.
McKinsey estimates that demand for data translators could reach anywhere between two and four million in the US alone over the next decade - constituting 20-40% of the 9.5 million US graduates in the business STEM fields expected over the same period. Demand is clearly going to be high, and companies are going to have to do something special to lure them in.
AI is undeniably a game changing technology that will affect everything, with pretty much no exceptions. However, it will not just happen. While companies will, understandably, want to adopt it ahead of their competitors, but it is not just a case of buying some cool new AI and hoping for the best. For organizations to realize the potential of AI, they need to have the necessary data structure in place. Machine learning - as is implied by the name - needs data to learn from in order to improve and work properly. AI supervisors will, therefore, be required to ensure that the right data is being fed in, and to feed in new rules, business logic, and feedback that will make it able to do the job. There has been much written about AI taking jobs, but in the immediate future, it will require humans to supervise the software, continuously teaching it and guiding it.
Low-quality data is a scourge with potentially dire consequences. Without clean data, any insights gleaned are next to worthless, and what could have been a great business tool becomes a potential weapon of mass destruction. Making decisions based on low-quality data means they will likely be wrong, while the time and resources wasted chasing out of date leads can be hugely detrimental.
It is also not a rare problem. According to Experian data, 78% of companies have trouble getting emails delivered because of inaccurate, 81% of retailers cannot leverage loyalty programs, and 87% of financial institutions have difficulty obtaining reliable intelligence. However, many are beginning to at last recognize that there is a problem, and we are increasingly seeing data hygienists introduced as a solution.
Data hygienists ensure that the data coming into the system is clean and accurate. They then monitor it throughout the entire data lifecycle. They scrub the data, dig out redundancies and duplicates - everything needed to clean it up ready for use. There are many vendors that promise software capable of doing this automatically, but such tools are simply unable to get data to the level of precision that employees with first-hand knowledge can by doing it manually.
The role of the data hygienist has been around for several years, yet uptake has been slow. Some 63% of companies still lack a coherent approach to data quality, according to Experian, and many data scientists have found workarounds or simply tried not to get too hung up on whether the data is clean enough or not. This year, we should see a renewed drive for data cleanliness and a surge in the hiring of data hygienists as a result. They should not be hard roles to fill. Data hygienists can be hired from administrative backgrounds and do not require any deep understanding of analytics or data, meaning they do not command especially high wages. They will, however, gain useful knowledge around data preparation from which to build a career in data, which should hold them in good stead and go some way to helping plug the data skills gap.