In a report published earlier this week, business consultancy Cap Gemini commissioned The Economist’s Intelligence Unit to find out what the leaders of 600 companies worldwide really thought about big data.

You can see an overview of the data on Cap Gemini’s website, which includes a nice infographic of the headline results, but it was one finding in particular that caught my eye. While 90% of the respondents said Big Data was an essential factor of production, 51% of them maintained that the single biggest factor that would prevent them from using their data to make better decisions was a lack of skilled people.

Skills are always a hot topic in technology, yet never more so in the case of a ‘hot’ trend like Big Data where demand for good people considerable outstrips supply. And even within the nascent Big Data industry there’s an ongoing debate over what kind of skills and abilities are most desirable. Until comparatively recently, generating commercial insight from large data sets was viewed as more of an art than a science and something entrusted to market research agencies or departments. As these large data sets became larger and analytic and Business Intelligence tools developed, these responsibilities migrated into a different part of the business often overseen by people with the word ‘analyst’ in their job titles. Now, however, as factors such as the growth of social media and the availability of sensor data have made data sets mindbogglingly huge, leveraging Big Data has become an end in itself for many organisations and something to be carried out by ‘data scientists’.

There’s not yet a fixed definition of what it is to be a data scientist. But it’s generally accepted that the role requires the kind of business insight you’d get from an MBA, plus a high degree of competence in statistics and quantitative analysis, computer science, natural language processing and machine learning. Given the sheer range of skills this role commands, it’s no surprise that the estimated average salary of a data scientist begins in the low six figures. It’s also a contributing factor as to why so many organisations choose to get the big software vendors on board when considering Big Data projects. After all, your average enterprise might find it very difficult to recruit a quant with an MBA who also manages to be a whizz at machine learning, but it’s reasonable to assume that an SAP or IBM can pool the relevant skills from across their large workforces.

Indeed it’s often the search for a critical mass of niche or hard to obtain skills that will compel organisations to outsource or contract out major IT projects. Big Data isn’t really any different. In the absence of single people who can do everything it makes sense to turn to a group of people who can combine their skills to the same effect. So when it comes to plugging the skills gap in Big Data, sensible organisations won’t be the ones who bite off more than they can chew. They’ll be the organisations who break the challenge down into manageable chunks.