An intro to data analytics projects and how to lay out your portfolio

Posted on: June 4, 2024

Post Category: Data

---

When you’re trying to land your first (entry-level or junior) data analyst job, it’s likely you don’t have too much relevant experience.

In this scenario, tailoring some of your past experiences to the requirements of the role can only get you so far… so when it comes to showcasing your value on your resume, how do you do it?

You create your own experiences.

For the lack of real-world experience you have on your resume, you can make up for it with personal projects or projects you built while taking a bootcamp or studying at Uni.

Now, for someone who’s just starting to build their data analyst career, it might be an attractive idea to put online courses on your resume, as proof of your skills…

But projects will have more weight.

Projects – like experience – showcase what you can build, what tools you can use, and the results you achieved/delivered.

So how do you start creating a project?

The standard workflow for a project is basically this:

• (You start with) The dataset and question
• (Then you do the) Analysis
• (And then you) Publish your work and/or a retro blog post

It’s a simple approach in concept, but a difficult one to execute when you are starting out (as most things are…)

And for beginners, it’s usually because of three things:

1. The lack of interesting beginner-friendly datasets/questions for aspiring analysts to tackle
2. The search to produce something fresh – or something “perfect”, and
3. Not having a system that helps them ideate a project, publish their work, and improve.

Let’s go through each of these:

1. Lack of interesting beginner-friendly datasets/questions

First of all, let’s define what I mean by an interesting dataset/question; when I say interesting, I mean something that can hold your attention while building your project out, not something that makes you stand out when you publish it.

Looking to create a project that makes you stand out – is not a realistic expectation for your first couple of projects, and I will talk more about this in the next section.

But back to the topic… your first couple of projects should generally be used for you to build your skills. And it helps to find datasets/questions that are interesting and appropriate for beginners to work through.

There are a couple of resources that you can refer to when building your first couple of projects:

• Maven Analytics (with their guided projects)
• MakeoverMonday (with their clean datasets and simple data visualisation exercises)
• Dataquest (with their guided projects), and
• Datacamp (with their guided projects)

Personally, I have not tried all of these resources (I’m only familiar with MakevoerMonday). But I know the value of having beginner-friendly projects to start with, and have heard good things about the other two.

2. The search to produce something fresh or “perfect”

It’s an appealing idea to produce something that is unique – unique to you, and something that will help you stand out amongst the other analysts.

But as mentioned, getting to that level takes time, practice and experience.

Instead, start by:

• (At least) Looking back at an online course, and trying to give yourself a more challenging task to stretch your skills, instead of just following along with the instructor
• Working on a guided project, or
• Making building and delivering a commitment – by taking part in competitions/hackathons, and submitting whatever you produce

There is absolutely no shame in doing something that is relatively easy, so long as your stretching yourself.

Building your portfolio can be challenging at the start. But once you build a few, you can start working on more interesting datasets, and you will have more experience (with the tool) to pull from.

Take myself as an example:

• My first couple of projects weren’t even mine – they’re duplicates of things shared in online courses
• My next couple of projects were from Uni competitions/courses and MakeoverMonday
• And it was only after those projects I had the courage to explore other techniques/topics to try. I analysed one of my favourite TV series, one of my favourite video games, graduate salaries, and the most popular coffee order/style at my work office.

It doesn’t have to be perfect; just keep practicing – and practice deliberately.

3. Not having a system that helps you ideate a project, publish your work, and improve

Ideating a project will be more of a concern once you get some experience with projects and start moving away from guided ones.

But regardless of whether you’re a beginner or not, I think you’ll get value out of the following content. I’ll also talk more about publishing your work and how to progressively improve your work.

When ideating and creating a project, I would focus on the following things:

• Ideate topics, questions and datasets. Write down a list of topics, questions and datasets that you are curious about. This can be anything, so long as they are interesting to you – because the more curious you are, the more likely you’ll do the project from start to finish. Keep in mind that you don’t need to execute on these ideas right away – the point is to create a list that you can refer back to whenever you have time/capacity to work on another project.
• Creating a user persona. The fundamental purpose of data analytics is to provide value to someone, and to provide value you need to understand the audience/user you are building for.
  • If you are not familiar, a user persona is a term used by designers. Before designers create an app/website, they think about the demographic/characteristics of who their user could be; these hypothetical users become their ‘user personas’.
  • Think about who *your* user persona is, given your dataset or question: “What is their demographic and their level of expertise and tech experience?”, and “What might they be concerned about?”.
  • Put yourself into their shoes and ask yourself these kinds of questions:
    • “What would they like to see?”
    • “What kind of problem should my analysis solve?”
    • “What kind of purpose should my analysis serve?”
    • “What should they get out of my analysis?”, and
    • “How should they react/interact with my analysis?”
  • Keep in mind that your user persona can be yourself – and you can create a valuable project *for* yourself.
• Plan your approach. Once you create a user persona, you should plan what the output should look like, and the tools you will be using to get there. But it’s important to think about *why* you are going to create your output in such a way – this is why it’s good to know your user persona.
  • For example: reporting KPIs to a business executive will look different to reporting those same KPIs to the business’s social media audience – one might require a dashboard, the other would need an infographic with big numbers.
  • Knowing how to articulate why you made certain design decisions (as above) helps you improve as an analyst – and it will be helpful when adding context to your personal projects, which I will talk more about in the next section.
  • If you want to go the extra distance with this step, I would also recommend applying visual best practices, so that you are also very intentional about your chart and formatting choices – you can read more about it here.
• Execute, create and publish. Execute on your plan, produce the desired output, publish your visualisations/code and share what you built with your online network.

When publishing your work, think about what platforms you’d like to use, and put your projects/content on there.

Platforms like GitHub, Tableau Public and novyPro are good enough platforms for your data analyst portfolio – and you can use platforms like Medium, or your own personal website, if you want to get around to writing blog posts.

And while you work through these projects, here’s what I would recommend so you can progressively improve your work

• Compare your output to those who are ahead of you in their journey. If you saw someone produce something stunning, or something that gained alot of attention, save their work so you can analyse and apply bits of what they are doing differently.
• Learn about best practices along the way. Read books on visual best practices and data storytelling so you can create projects that look good and that people might want to use. Tableau Public’s Viz of The Day is a good place to find examples of this.
• Make your previous work easy to reference. Save all your project work in the same folder(s), so you can reference them later when you build a project with similar steps. This will save you time and allow you to dedicate attention to newer techniques.

How to lay out your portfolio

In my experience, to date, I’ve used two platforms to publish my personal projects: Tableau Public and GitHub. For all of my Tableau data-visualisation projects, I use Tableau Public. And for all my code-related projects, I use GitHub.

But Tableau Public can be easily be replaced by a more generalised platform for data visualisation like novyPro, which helps you publish projects in Tableau, Power BI *and* Excel.

Regardless, these are probably the only platforms you would need for you portfolio: one for code, one for no-code projects. Blogs and personal websites can be added at a later stage.

Here are some things you can do to make your portfolio easy to navigate – for your own reference, and for any recruiters that would look into your work:

Now, to be honest, I do not follow these practices often – especially since starting my first analyst role. But in the (very) rare scenario that a recruiter looks at your resume and wants to see/interact with one of your projects, it’s good to provide at least a little bit of context…

It’s a good habit to do right after you finish a project, since you have the grunt-work still fresh in your mind. But I’d say this bit of extra work is a nice-to-have, and not essential.

For GitHub:

Use the readme file to summarise your work:

• Write a brief and informative summary of your project. This can be one or two sentences, highlighting what you produced and what purpose it serves i.e., what value it provides for its intended (hypothetical) user, or what output it supports.
• Write a bulleted list of modules and/or functions you have used to support your analysis. For SQL, focus on listing functions. For programming languages like Python and R, focus on packages/modules. In case a hiring manager wants to look at your project, this list will give them a good idea of the code you can build.

For Tableau Public and/or novyPro:

Use the description field to summarise your work:

• Write a brief and informative summary of your project. This can be one or two sentences, highlighting what you produced and what purpose it serves i.e., what value it provides for its intended (hypothetical) user, or what output it supports.
• Write a short list of techniques you used/explored to support your output. Reserve the list for up to 5 key techniques you used. If you created the project to specifically explore new techniques/features of the app, make sure they are included in the list.
• (Optional) Mention your sources of inspiration. Mention any data visualisations that have inspired your work – and credit the authors. List any resources (YouTube videos, blog posts, etc.) that gave you inspiration and support.

Now, for blog posts… I’d say it’s optional, but (again) it’s nice-to-have. You don’t have to follow specific guidelines (besides delivering on the title of your post), and you can write about anything – it can be a tutorial, a breakdown of a personal project, your favourite resources, etc.

But there is one key benefit – which Annie Nelson highlights well in her book ‘How to Become a Data Analyst’. And that key benefit is being able to articulate your knowledge and the analysis/thought process behind the projects you build.

Annie mentions that, when you apply for Data Analyst roles, you may come across technical/panel interviews – and during these types of interviews, you may be asked to showcase a project you built. And when you’ve written a blog post on it, you can talk about it with ease.

I personally never had to do a technical interview like this, but I know of people who have.

Closing remarks

For any job you apply for, experience is looked highly upon.

But many aspiring data analysts do not have (enough) experience.

That’s why aspiring data analysts require a portfolio of projects to make up for it.

If there’s only one next-action you could take away from this post…

It’d be to set up GitHub and Tableau Public or novyPro, and start looking at some resources for beginner-friendly project datasets.

Create your own experiences, when you don’t have many, to aid your analyst job search.