How to avoid the rabbit/duck holes
How to avoid the rabbit holes…
Over the last couple of years, all things ‘big data’ related have become established as the leitmotif in analytical circles. Drawing out actionable insights from a wealth of available customer behaviour, channel and product data is a challenge that many organisations, across all retail industries, are facing into. The opportunity for analysts to identify new learnings - and the associated anticipation from business teams - is great, but the risks of getting lost down multiple rabbit-holes of impenetrable information and becoming overwhelmed by volumes of irrelevant data are all too apparent.
…is it more science than Art?
Whilst several technology platforms and software tools are readily available to support discovery analytics and big data explorations, having appropriately skilled and engaged people is absolutely key. The rise of the ‘data scientist’ has challenged the traditional understanding of a ‘data analyst’. Various descriptions of a data scientist exist – one of the neatest articulations (cribbed from various sources) suggests a data scientist is someone who is better at statistics than any software engineer and better at software engineering than any statistician.
One scientist role to rule them all?
What do you think the blend of knowledge, skills and behaviour in a data scientist role actually looks like? And are those attributes best embodied within one individual, or spread over a number of specialised roles that work closely together? Do the activities of data acquisition, preparation and manipulation (and associated technical aspects of data wrangling or munging) sit neatly alongside the investigative, and often iterative, analysis and exploration required to generate new insights? Does the data scientist role also need to possess the visualisation and presentation skills to effectively communicate new insights with (a typically less analytical, and perhaps numerate) colleagues around the business?
Please come join the Group: https://www.linkedin.com/groups/8452074
Good article. Too often these terms are used interchangeably but I think your combination of software engineer and stats is a good starting point. The more advanced DS guys need a bit of machine learning, customer experience, process design, org strategy, marketing, design thinking, etc. Would love to rejoin the team. Can you open an Australian branch? 😀
Positively provocative with a good dose of reality