2 ways to organise a Data Team

2 Ways to Organise a Data team

The terms Data scientist and data engineer and analyst are ubiquitous within the Data sector now and often used interchangeably. Some people might think inaccurately but that is because the roles and job titles can be so varied.

This is clearly a generalisation but broadly speaking there are two ways to set up a data team.

Firstly though it is imperative to understand that all organisations need one thing from their Data before anything else and therefore it is the starting point for all Data leaders.

 Clean Data

You get clean data by having a consistency of data collection and storage but it will still need further work even if you have this. Most organisations don’t have this, to begin with and even if they do you are then combining with other data.

Organising a data team: Option one

You have a group of people who clean and store the data using a process & method that it is easily accessible and consistent for others to work on.

  • In this case, you want those Data people to a broad range of knowledge & experience with the types of data you work with
  • be highly proficient at coding probably familiar with cloud storage and big data or unstructured databases and infrastructures

The second group of people will likely have skills in modelling or statistics if you have need of predictive models or visualisation and data storytelling experience to explain the picture to other less technical stakeholders – skills that will be important are

  • ability to access data from a Data warehouse or lake
  • some coding & statistical modelling.
  • soft skills in presentation and communication
  • as well as a more business-oriented focus.

 

Organising a Data team: Option two

You don’t want two distinct sets of skills you have a group of people who both work on cleaning the data and modelling it.

Here you will still likely want all the same skills as in the previous option but combined – not all of the people will necessarily have all of the skills but you likely still need them within the team somewhere

 

  • team members who have a knowledge of coding, statistical modelling and data storage and infrastructure
  • team members who can code, build production models and visualise data
  • team members who can access data & visualise data and approach problems from a business needs point of view

 

The second model offers more flexibility than the first but also asks for the combination of more different skills.

This is not an exhaustive list but a basic catch-all that describes two distinct ways in which data teams tend to be built most organisations will have teams that fall into one or other.

 

To read more about Wyatt Partners services