In this blog I will go over my experiences and thoughts about being the project lead for a client project at the Data School. As I don’t want any lawsuits that will probably make this my first and last project as team lead. I will refrain from mentioning any specific details. This will just be a brief chronological overview along with some things I think I did well or I can improve on.

What is a Client Week?

A client week is part of the Data School training initiative. It involves corporations giving those in the Data School some of their real industry problems and a little bit of their time in exchange for free and more often than not valuable insights and perspectives. It is often a high pressure task as often you have to plan, clean the data, understand the data, create insights using Tableau or Alteryx and then finally prepare and present it.


Requirement Gathering



In the requirements gathering meeting, the client gave a brief overview of the entire project, as well as what format and versions of Tableau and Alteryx they used. Being my first requirement gathering as the team leader, I wasn’t sure in what to take notes of. So I opted to take notes of almost everything the client said, this resulted in a stack of pages. Most of those pages never saw any use. I also asked the client for a project overview in a written form that describes the business problems and also what they have done previously.



Asking the client for a written brief/presentation:

This is especially important as in previous experiences, with no presentable brief, there is often a lack of clarity. An example would be our second project where there had to be plenty of mid-week changes and tears when certain definitions and goals were interpreted wrongly.

Software details:

Overlooking what versions of the software and understanding which platform to use can actually be a huge detriment. Presenting good information or software only for it to be unusable can turn a successful project into a useless effort.



Breakdown the notes:

I recorded way too many notes. It only served to alter on clutter the valuable information I had in there. Next time I would definitely try to add headers on my notes and separate everything. Like breaking it down to the business problems, issues they have had etc.





As the client didn’t give us any data beforehand, I spent the weekend snacking and watching Netflix. I went in with slightly more knowledge of what to come than my team. Their brief and presentation was quite well explained and the client took time to answer all our questions. After having a quick meeting discussing how to address each business problem, we started working on the project immediately. As the client wanted extra data, I started by telling everyone to research where we could acquire reliable but most important free data.

After mindlessly researching through data sources (which thankfully still worked out), I finally realised to tell everyone to collect their results in a spreadsheet. Only about two people did. We then went through the sources and decided on our roles and which sources to use. Smaller mini groups were formed and the project initiation was complete.



Identifying how to address business problems quickly:

Spending too long on trying to address the business problems early on is a draining process for the team members. Having just been through a brief, anymore words would go through one ear and fly out the other. The methods also might change as the project goes on so I think discussing it on a high level was a good idea.

Breaking into groups quickly:

Assigning the task quickly and braking into mini-groups was also a great idea. For these types of shorter project I think a breadth approach where we cover a wide range of topics briefly is great way to showcase what’s possible. I think it also inspires creativity and I would often allow the members to go through with their ideas as long as I think it will bring value. If they actually want to do something, their productivity will go up.



More preparation:

Even though I didn’t receive any data early from the client. I should have still started by at least attempting the first few task by myself. Having a try on these tasks even on a very high level would give me a better understanding of what instructions to give people.

Clearer Instructions:

I simply stated to research. This wasn’t because I wanted to give vague instructions but rather that I was unclear what to research myself either. This leads in from the previous point. I also should have stated to put it in an excel earlier to organise all the data sources neatly as after people finished, they often don’t want to revisit the same site.

The Data School
Author: The Data School