Navigating Your Data Learning Journey: Key Insights — Tip #2

Photo by Ugne Vasyliute on Unsplash

Starting with curiosity is a good entrance to the data world, check out #1 Start with Curiosity. Now, with so many skills, tools and professions, how do you make the choice?

Spreadsheet as a Launchpad

Spreadsheet may be the most common tool comes in people’s mind when manipulating data. It’s easy to obtain and use. Most people know how to do simple formula, e.g. SUM(), AVERAGE(), analysis techniques, e.g. add a pivot table, and simple charts, e.g. Pie chart.

If you know how to use a spreadsheet, you’re well on your way to becoming a competent data practitioner.

Even more, a friend always says: If someone is good at spreadsheet, s/he can do coding well. When he consults a company, he always ask the company to recommend an employee who is good at spreadsheets, then trends the employee to maintain the data project after his consultation is finished.

When you do play these numbers in a spreadsheet, you’ll learn what are dimensions and metrics, SUM, AVERAGE, MIN, MAX, etc, these fundamental data knowledge. These may be your unknown-konwn: you might not realize that you already process basic data knowledge 🤩. When I recruited a summer intern a few years ago, I asked them about it. 2 out of 20+ sophomore or junior college students could explain what dimensions and metrics are.

Charting Your Learning Path by Your Problem

I started with merely hundreds of rows and gradually escalated to thousands. At this point, my only familiar tool was a spreadsheet. But as I explored more, I stumbled upon a fantastic resource, BenCollins, that helped me elevate my spreadsheet skills and introduced me to App Script.

My task was to comprehend the monetization results from over 10 ad networks:

1. Fetching reports from ad networks and importing them into my spreadsheet.
2. Mapping this data to ad placements for future ease of analysis.
3. Converting the bulky 100K+ rows of data into concise reports or charts for stakeholder presentation.

Spreadsheet played important role in my problem but not a complete solution. The repetitive nature of these steps was a time drain, eating up 30% of my daily hours 😣.

Let me introduce Known-Unknown Matrix with my example:

Known-Unknown Matrix

Unless I explored outside of my Known, I wouldn’t have discovered App Script. The discovery moved it from Unknown-Uknown to Known-Unknown — something I was aware of but hadn’t yet mastered.

Then I applied it on my problem, I found it wasn’t an ideal solution thus moving it from Known-Unknown to Known-Known.

Now, think about your problem. What challenge do you aim to overcome?

Seek out new tools and approaches, and experiment with them on your problem. Embrace the ethos of learning by doing.

That’s how I came across Python and SQL. These were the exact tools I needed to solve my problem.

Start Small and Use Free Resources

Google is a great resource for research, but you need to know the keyword. You can start from the keywords you know, then you’ll find new keywords showing in the search result. That’s the way to uncover your unknown.

There are plenty good and free resources,

• YouTube, use it as Google.
• Help Page of the tools you used. There must be explanation for glossary or key concept.
• Ask people. Describe your problem to your colleagues, team members or friends. Get inspiration.
• ChatGPT. Describe your problem to AI.
• Join webinar, online events of the tools you used. At the beginning, you may not find useful or don’t understand what these speakers are talking. However, you’ll get a direction. Or at least you know this is not what you need.

When you get direction, say you found Python or SQL. There are also many good and free resources,

• SQL introduction by W3C
• Google Skillshop for BigQuery
• You can run SQL in BigQuery. Each session is 30min ~ 1 hour.
• There are public dataset you can use.

Always back to your problem. What things you want to do? What issues you have when you do these things?

Google your problem. Then you’ll find out how to state a good question. When you do, you dive deeper and move unknown-unknown to known-unknown.

When To Invest in Structured Learning

After you accomplish many small tasks, moving several dots into known-unkown or known-known, you might still feel you lack of an overview.

If you have some time and are not too pressed to deliver results, you have time for trial and error in the process.

That’s a good timing to take a course to enhance your knowledge. Again, there are plenty short or trial course you should get a taste,

• Coursera provides some free courses.
• Python for Everyone
• DataCamp provides trail.

It’s important to apply your learning while you are learning. Doing the assessments or assignments are not enough. To apply your learning to the work you need to delivery increases your motivation and desire to explore.

More to come..

In this post, I encourage you to step in the data world to resolve your problem. Keep your curiosity and motivation will help you become a lifelong learner. In the next post, I’ll talk about common problems you might encounter after taking the first step.

🤩 I’m happy to hear how do you build product, what data do you check. Feel free to reach out to me on LinkedIn Karen Hsieh or Twitter @ijac_wei.

🙋🙋‍♀️ Welcome to Ask Me Anything.