Over the years I have had to select many new staff into the highly specialist world of data analysis. Not only does the new analyst have to fit the business team, they have to be capable of wrangling “real” data.

Within some organisations, skills testing is routine for new candidates but it is often a standard assessment across the business.

So how do you know your analyst is exactly that, an analyst?

Try asking an off the wall question: How many clocks are there in the UK? (Can be substituted for any country).

This is where we sort the analysts from the Excel users.

My preference is to leave the question hanging with no further explanation, maybe some friendly encouragement but not an explanation of what you want. I have received a multitude of different answers and please remember, there is no wrong answer.

One answer I received was "There is only one, Greenwich Mean Time."

Whilst very logical, it's an easy answer.

What a good analyst would consider is the following:

Every town has a town clock.

Every house has at least ‘x’ clocks, a clock on the cooker, an alarm clock, a wall clock.

So now I start to push for an answer, “so how many would that be in the UK?”

What I look for is a good estimation with some accurate mental arithmetic. I don’t mind if the candidate has to guess 60 million people and 30 million houses so long as 5 clocks per house and 1 watch per person doesn’t add up to 2.1 billion clocks.

The analyst thinks about scenarios, makes assumptions, qualifies their answers and has good dead reckoning skills. This should not be underestimated. It is the skill that makes an analyst recognise when a result just doesn't look right.

Going back to my first point, so how do you know your analyst is exactly that, an analyst?

Tailored candidate assessments can save you a lot of time, resources and money.

Kevin Reid

Subscribe for updates