So we talked about Data Science previously so lets get this sorted. First of all there are multiple ways to get into Data Science for your benefit I am going to outline them here. Business, Hadoop (Big Data), Mathematics, Statistics and programming.
The profile of how you entered into Data Science is very important to where you need to be in the organization. Remember while you are essentially serving the business you are also knee deep in data (and really all kinds of data at that). Hopefully someone has put together a nice data lake somewhere for you to work successfully in and you do not have to spend your days doing all the data wrangling (surveys say that approximately 70%) of your time is spent doing this (and that is a waste). So now some interesting data points from some of my interviews based on sizes of data sets becomes relevant. The size of the data sets handled related back to how you came into Data Science (not always but based on a sample size of 79 data scientists) which makes sense if you think about it, most people that have come from Hadoop or Big Data initiatives are much more used to handling large scale than someone coming from the Business as a power user using Excel. While this may not always be relevant when you are training a model and managing scale it definitely is and so keep this in mind when training models that need a lot of data where your data scientist entered the fray. Also consider looking at changing some roles up a bit, sometimes you do not need a doctor you need a Physicians assistant (lower cost but skill set still relevant to the job) and thus in Data Science you can look at them as Data Business People, Data Citizens, Data Curators (still debating this one but hang in there). The key to data usage is to spread it through the organization and make sure that people are speaking that language. Often a Data business person will be more likely to spread this understanding of data (data speak) whereas someone with a more focused Data Science role may not. Part of it is their mobility in a matrixed organization and the ability to translate the abstract to the relevant. So to sum up you need to think about the profile of the Data Person you are working with and how that helps your organization grow. Also I mentioned data curation, while not a direct data science mandate this one is a big deal. The most often cited road block we see at clients is the fact that the data is not available when it is needed. A Data business person can help alleviate by working in the gray area and making sure that the prep work is done by someone. This means you ask a question you get an answer and then you know you are on your way.