In which I ghost-write part 2 of Avinash's post

In my last post, I pointed you to Avinash's post about Reporting Squirrels versus Analysis Ninjas. My focus in that post is the underlying concern of "return on analytics" or lack thereof. This post takes up Avinash's argument directly. Think of this as part 2 of Avinash's post if he had kept on writing.

Avinash's summary of his post is as follows:

Reporting Squirrel type work has a minor incremental impact on a company's bottom-line, hence your career progression will be on a slower track. Analysis Ninja work on the other hand does (even though their output is not numbers, their output is words in English!), so you want to be on that track. Seek those opportunities.

In short, Reporting Squirrels are the people who work on data processing, loading, transformation, and reporting while Analysis Ninjas are creative problem solvers who work with the cleaned up data.

***

One approach to this issue is to understand the Analytics Process. Raw data is never the right answer (see Chapter 6 of Numbersense for more (link)) The Analytics process starts with data processing, loading, transformation, and reporting, and on top of a foundation of good data, one can do slicing, dicing, modeling, etc.

In any practical Analytics process, 70 to 80 percent of the effort is spent on processing and cleaning data! Those are "low" value-add activities when considered alone. But these are not dispensable tasks either, and somebody has to do them. Too bad textbooks and academic courses tend to spend less than 10 percent of the time on data processing and cleaning, and thus contribute to the skewed view of the analytics work.

If you are managing an analytics team, you have a decision to make. Do you create a Department of Squirrels and a Department of Ninjas? Or do you create one Department of half-Squirrel-half-Ninjas? I have usually preferred the latter which avoids the caste system but others have set up separate departments. The reason why I like to combine the two roles is that the best Ninjas must have a complete grasp of the data, including every assumption and every adjustment that was made or attempted (and failed).

***

Avinash has advice for people getting into this field - get on the Analysis Ninja track.

That is good advice, with a caveat. The two roles have different requirements, and not everyone is suited for both. Reporting Squirrels need to be disciplined, care about order, structure, proper process and documentation, and attention to the smallest details. and not easily irritated by anomalies. Analysis Ninjas have to be creative, adaptable, able to think outside the box, and have great communications skills and business sense. The overlap between the two sets of requirements is surprisingly small.

For job-seekers, you should assess your own temperament and strengths. When you apply for a job, figure out from the hiring manager what types of people he's looking for. Understand how the manager organizes the Ninjas and Squirrels, and figure out where you fit in.

If you think along these lines, you might notice that my previous advice to create unified roles also comes with a problem: it is hard to find in one person who is an A in both roles; most likely, you'll find someone who's A in one role and B in the other, and you'd have to work around that.

 

 

Comments

[Hitesh Arora]

Hi Kaiser

I totally agree with you that no matter how much you want or try, you can't really separate Ninja from Squirrel. Avinash's perspective is more relevant to large scale companies where there are enough Analysts for both type of work but in mid size to any average company, it is not so as they are still evolving. Personally i feel more confident about giving insights, if i thoroughly check and slice the data myself

Regards
Hitesh