A few week’s back I wrote about the data translator – a role on the data science team that bridges the gap between the data scientist and the business.
That got me thinking – what are the characteristics (that may be less obvious) that make for a good data analyst?
1. Intellectual curiosity and focus
I often hear people say “data is boring”.
Intellectual curiosity is that character trait that allows the good data analyst to find the interesting in the data ( and it is almost always there.)
In some cases, the interesting may be the similarity between this data set, and other similar data sets that the analyst has seen before. In some cases, the interesting may be the variations or anomalies shown between data sets.
Or it may be some small, almost irrelevant detail in some attribute.
Most critically, interest is driven by the ability to link what you are seeing in the data to the data’s fitness for purpose. In other words, that good analyst is inherently curious to understand whether the data is of sufficient quality to solve the business problems for which it is intended.
Yet, data is vast. The intellectually curious analyst can easily be side tracked by the various anomalies, curiosities and other false trails that may divert him, or her, from his purpose.
The best analysts are able to focus their attention on those issues and trails that lead to solutions for the problem on which they are working.
Other issues may be noticed, and marked for future attention, but they will not overwhelm.
2. Situational awareness
A term most commonly associated with pilots, situational awareness is the ability to know where you are in relation to the ground, and other obstacles, even when presented with complex problems.
A lack of situational awareness often results in pilots flying into the ground
For analysts, I use this characteristic to describe the ability to maintain a big picture view of the problem, and related sub issues, even as one gets into the detail.
Data analysis may require one to explore multiple paths in order to answer a question.
One also needs to be able to document one’s findings in a coherent fashion in order to provide the reader with the detail they need to come to a decision.
3. The ability to second guess one’s self
The best analysts recognise that their conclusions may be driven by bias, or by a lack of understanding of the underlying business problem.
In other words, what may appear to be an issues based on the data may in fact be irrelevant to the business. Rather than couching a discovery as an issue, the best analysts will refer the finding to other members of the team and use their input to either sharpen focus, or move elsewhere.
In many cases, as an external eye, I find potential issues that are simply not a focus now.
Rather than hammer these issues, it is important to get validation from business that these are in fact issues, and be prepared to park them for later if no one else is concerned.
4. People skills
While data analysts are often thought of a tech geeks ( and some are), the best analysts recognise that analysis and problem solving must combine their insights with those of others.
The analyst must work with various stakeholders to understand the business problem and hone his theories before diving too deep into the data, must have the ability to test his findings with business and technical stakeholders, and must manage expectations around both what is found (or not found) and what must be done to address concerns.
Tact and consideration are important skills to ensure that issues are addressed
5. Story telling
Last but not least, the best analysts can take their findings and present them as a story that compels decision makers to act. This ability ties the various abilities above into a coherent whole.
The story must provide the reader with context – so that they can understand both the extent of the problem and, ideally, the business impact – as well as the detail of underlying issues.
What do you think?
What are the most critical attributes of great data analysts?