Leo Paas: Coming down from the ivory tower

Last month's 'PwC Herald Talks: Big Data' event has been the catalyst for an ongoing conversation on LinkedIn and other places about the need for business-savvy data scientists.

As the leader of Massey University's business analytics programme, a number of attendees reached out afterwards to ask what the tertiary education sector can do to help develop this special combination of skills, and to quickly produce graduates to fill the skills shortage in this area.

As Rachel Harrison from Data Insight pointed out at the event, technical data skills are only part of what is required; the ability to understand business issues and communicate findings is the other crucial part of the equation.

I'll be the first to admit that academics often like being in their ivory towers but there is a real push by today's business schools to connect with the business community and things are starting to change. It's no longer good enough for largely government-funded researchers to sit on their technical skills and not get their hands dirty in the real world of business. Having a "great statistics department" isn't an end in itself.

Universities need to mobilise their statistics departments to produce graduates that are capable of tackling the challenges that businesses face and identifying the opportunities that will create success at an individual, firm and country level.

But educating students is a long process. I can confidently say that data scientists with buisness nous are going to emerge from tertiary programmes - but it's not going to happen overnight.

Businesses prefer to see immediate results but, given the well-documented shortage of good data analysts, it is surely better to work on developing a talent pipeline for the future than to do nothing?

I would like to open a discussion with businesses about investing time in a postgraduate student. Massey's master's programme requires its students to complete a real data analytics project. To do this, they must understand an organisation's objectives and data sources. They put into practice what they are learning through their studies, with experienced academics overseeing their work and providing advice.

For those businesses willing to take a longer-term view to staff recruitment and project development, I can assure you that we are not just "teaching stats". Everything the students learn is embedded in a framework that aims to answer real questions for businesses.

Some examples of this work include attribution modelling for companies to identify which of their advertising channels is producing the most effective results and analysing the profitability of loyalty programmes. These are practical projects that deliver useful insights to the organisations the students are woking for.

The other important outcome from these sorts of internships is they help to bridge the gap between businesses and universities. The flow of information between the two is increased so research is directed towards the right areas and educational programmes produce graduates with the right skills.

More collaboration between businesses and academia is crucial. While universities can't solve the big data skills shortage in the short term, we can at least soften the problem in the longer term by working together.

If businesses are willing to shift their thinking to a longer-term horizon, there are advantages to partnering with a university to help develop a postgraduate student.

At the end of the process there's every chance they will have a data analyst who already understands their business. In an environment where vacancies are already hard to fill, that sounds like a pretty good potential employee to me.