Data science is on the rise. A couple of years back Harvard Business Review suggested that ‘data scientist’ is the sexiest job title of the twenty first century and the hype around data science shows no sign of abating. The term ‘data scientist’ itself was only coined in 2008 but since then the number of data science roles in organisations has grown exponentially, as the volume of data available for analysis also grows. But this presents a challenge for organisations – in such a new and fast-changing field how can they identify the skills they need, find appropriate people who have those skills, attract them into the organisation and then retain them in the long term?
A good data scientist (or analyst, as we used to call them!) is absolutely vital in any analytics project. At Smart Vision I’ve worked in support of hundreds of projects in all kinds of industries and applications (predicting tax fraud, predicting which parents to target with a specific Christmas toy, determining the best path of training to get a long term unemployed individual back into work, predicting what fire safety equipment to sell into particular sites, grouping stores by type to determine stock types, predicting newspaper sales by day and by shop to maximise sales without waste, predicting what lingerie will sell in what combinations to certain women….the list goes on and on and on) and what I’ve learned is that, as sure as eggs is eggs, they all stand or fall on the skills of the analyst.
In-house data science skills are critical
But not just any analyst. An in house analyst. Having someone inside an organisation in a dedicated role is ALWAYS more effective than using an external resource. ALWAYS. At Smart Vision we have plenty of experience in training analysts and working side by side with them across their first few projects to get them started and we see this time and again. If you want your analytics project to succeed you really need to build the analytics skills of your in house team. It’s tempting to outsource the analytics heavy lifting to external analysts and consultants. They can certainly deliver the analytics capabilities but they’ll never really, truly understand your business in the way that an internal person can and it’s this business understanding that is the key to analytics success. The business-focused, data-literate people you already have can be taught advanced analytics and modelling, but it’s next to impossible to embed external analysts in the culture and history of your organisation.
It’s unusual that all the weight of an analytics project falls on the sholders of one person. More likely, you’ll have a team of people working on your data analysis projects together. Having a skilled analyst is not the only thing you need in order to make a project successful. So what are the key skills required and where in your organisation are you most likely to find them? In my view the skills required can be split into four broad areas and the chances are that you’ll already have most of what you need within your organisation:
1. Business and organisation-specific knowledge
This is vital. Successful analytics projects depend on a high level of organisation-specific knowledge and understanding of the type that your customer-facing staff are likely to have. People who regularly work with customers – sales people, customer service agents, account handlers, service delivery staff – generally have the greatest breadth and depth of organisation-specific expertise. They are the people who understand how the business really works and what it is that your customers most want. Their involvement should be central to all analytical, predictive and prescriptive process that are developed and deployed.
2. Analytical expertise including a grounding in analytical process development and analytical methodology
You may not have this specific skill set already in your organisation. External expertise and training can help you to ensure that your in-house personnel are exposed to and learn how to apply advanced analytics and modelling techniques to support prescriptive analytics applications. As I’ve already discussed, it’s much better to train up the business-focused, data-literate people you already have and give them the advanced analytics and modelling skills they need than to outsource this whole function, tempting as that might seem.
3. Access to relevant data access and the local knowledge necessary to interpret it
Data is the raw material of the entire predictive and prescriptive analytical approach. Again, it’s your existing staff who will have the necessary expertise in the interpretation and use of the wide array of data required. A data audit, involving staff from all areas of the business, often reveals that you have much more data already available to you than you might thing. Again, external consultants can’t possibly know about all the data sources you already have, or understand how the data was collected and the implications that might have for its use.
4. Expertise using predictive analytics technology
You need people who are experienced at using some form of predictive analytics technology, and it needs to be something that has been developed specifically for the purpose of predictive analytics. You can’t do this with Excel. If you don’t already have an analytics tools such as IBM SPSS Statistics or Modeler then you’ll need to invest in one. Here again external consultants can advise on what tools you need and offer training and support to get you up and running. At Smart Vision Europe we’re highly experienced in supporting the implementation of IBM prescriptive analytical technology in a wide range of organisations, including the training and development of in-house analytical teams.
Don’t lose sight of all of these skills when embarking on the project – a single data scientist hiring into your organisation will not have this full range immediately. They may never develop the full range because of a lack of tenure or siloed expertise in your organisation. Involving people from all aspects of the business in combination with external advisers where appropriate will ensure your expensive and highly skilled analyst gets the very best opportunity to make a competitive difference in your organisation.
