• Infinite Upside
  • Posts
  • IU #002: Feeling Overwhelmed About Creating a Data Portfolio?

IU #002: Feeling Overwhelmed About Creating a Data Portfolio?

Here are 11 tips to help you get started

Author

Michael Dillon
April 21, 2023

Welcome to the second edition of Infinite Upside.

Here, I share how you can land a job through online networking.

Today I’ll explain an essential part of this strategy.

Your data portfolio.

I love asking my online network for help. So, I asked a few friends to share some tips with you.

One last thing before we begin.

If you didn’t receive this post in your emails, hit subscribe so you don’t miss it next Friday.

Let’s get started.

Read time: 5 mins.

What is a data portfolio?

Firstly, a data project is when you analyse a dataset to generate insights.

So, your data portfolio is the collection of your best data projects.

Why you should create a data portfolio.

There are so many benefits.

You get to learn new skills. If you are missing work experience, you can use a portfolio instead.

Businesses hate taking risks.

They hate it.

And the hiring process is a massive gamble for everyone involved.

Candidates and hiring managers are strangers to each other. People get jobs based on stories they tell in interviews.

The whole process is built on trust.

Share your data portfolio. 

Don’t SAY what you can do. SHOW them.

A great portfolio makes it much easier for a business to take a chance on you.

You become a sure thing. Now, it’s everyone else that seems like a gamble.

How to create your first data project

Here is a strategy to create your first data project. I have sprinkled their advice throughout.

1) Get started

I know what you are thinking, 'I'll do projects after I....'

STOP.

That perfect moment when you feel ready to start will never come. Don’t wait. Get started today.

‘You might not have the skills to complete the project, but that's fine. The beauty of data projects is that you'll learn those skills along the way.’

‘Get started and you will learn one of the most important skills in data. Googling.’

‘My first "real" data project was analyzing the database of sales of a pizzeria provided by Maven Analytics. It’s my personal website (via Carrd) which is linked to Maven.’

2) Choose a topic

Which industry do you want to work in? 'Marketing'

Which skill is missing on your resume? 'Python'

Use Python to do a marketing project. Simple as that. Your data projects can plug the gaps in your resume.

3) Find a dataset

There are many sources to find free datasets. I spoke with many data analysts and they gave me great ideas:

‘I recommend Maven Analytics free data playground. They will host your data portfolio for free, and give you sample case studies to pair with the data.’

‘I have just started creating my portfolio. I used Maven Analytics to host my portfolio and I used their datasets too.’

‘Kaggle is a great source for datasets. I created my portfolio using WIX. I like it because it looks professional and was easy to create. It does not have any advertisements.’

‘My first Data project was a SQL/Tableau project based on the show, Game of Thrones. I didn't know where to find datasets, so I just created my own. Now, I would recommend Maven Analytics.

“Find a dataset that interests you and aligns with your skill level. Choose something that you are interested in.'

‘My first project was on Amazon best-selling books. I analyzed customer behaviour & sales data. I found the dataset on Kaggle & hosted it on GitHub.’

‘I personally prefer Kaggle when I need data for projects. Since I am interested in sports, one of my first projects involved statistics of Cricket players.’

‘Alex Freberg’s YouTube videos helped me understand SQL projects. After completing the SQL part, I used the same dataset for creating the BI report too.’

‘My first project was SQL. I think it was on international breweries. I used PostgreSQL for this. The next was Excel and so on. I hosted some on my GitHub account.’

4) Define your hypothesis

Create a project plan. Before you start working, outline your goals, objectives, and timeline.

With data, it's possible to do anything. But you can't do everything. Focus is key.

Let’s pretend you are using Python for a marketing project.

State a hypothesis to investigate.

For example, 'There may be a correlation between email frequency & conversion rates. I predict that as we send more emails, the conversion rate will increase.'

5) Clean the data

If you do want to clean the data:

  • Check for missing values, duplicates and null values.

  • Make sure the dates are consistent.

  • Ensure the order of the email campaigns makes sense.

With data projects, people won't see the hours you spent cleaning the data. So, for your first projects, I recommend using clean datasets. Spend your time on the analysis & insights.

6) Explore the data

Use your Python skills to find meaningful insights.

  • Find the mean, median and standard deviations of conversion rates.

  • Visualise the data to spot outliers.

  • Split your email audience into different segments.

  • Does each audience segment respond differently to emails?

Be curious and play around with the data.

7) Transform the data

Calculate the conversion rates for each email campaign. Group them by date.

8) Test the hypothesis

Apply regression analysis to identify any patterns. Analyse email frequency and conversion rates.

Is there a relationship between these variables? Draw conclusions about the relationship.

9) Add business context

It's not about fancy code. It's about using this info to explain how a business should change its strategy.

These insights are how you show a business how you think. By offering solutions and explaining how your insights add value, you’re demonstrating how YOU add value.

'How are you deriving insights from your project in a way that others, using the same data, missed out on? Make sure that your projects involve specific, actionable recommendations that YOU found in the data.’

10) Share your work

There are many websites to host your portfolios.

Share screenshots of your work. Provide a summary of your insights on LinkedIn.

Be clear and concise. People like to skim. Add a link so people can dive deeper into your full work.

‘'Don't be afraid to ask for help or guidance.'

'I used Carrd to host my portfolio. It was free and not too difficult.'

‘Ask your peers to look over your work. They will catch mistakes & offer fresh insights.’

11) Keep going

Do each of these 10 steps consistently until you have built up a portfolio. This will help get you noticed by the right people.

“Your first data project is just for practice. Keep going”

‘Try reading notebooks on the same topics on Kaggle for some inspiration. It also helps in brainstorming questions to answer.’

In summary:

  1. Get started

  2. Choose a topic

  3. Find a dataset

  4. Define your hypothesis

  5. Clean the data

  6. Explore the data

  7. Transform the data

  8. Test the hypothesis

  9. Add business context

  10. Share your work

  11. Keep going

Thank you for the amazing support.

Be sure to tell your friends about Infinite Upside. I’d love to meet them too.

See you next week!

P.S.

If you want to learn how 64 data professionals landed their jobs in data, I have written an e-book, ‘How to Find a Job in Data Analytics.’

I asked them HOW they landed their amazing jobs, so you can land your dream job.

Thank you so much,

Michael