Self-Learning Data Analytics as a UX Designer: #2 What to Do When Starting Tracking from Scratch

From the last chapter, I learned the importance of collecting information upfront before diving into data analytics.

P.S. If you missed the 1st Chapter, explore here!

Self-learning about data analytics as a UX designer #1

Now is the time to get hands dirty with data tracking.

3 Approaches before starting tracking:

1. Understand your product.
2. Set up hypotheses, which in this case, I will apply from my exploration in the 1st chapter.
3. Select the approach to tracking.

Understand your product

This refers to how the flow works or how users will use your product. There is a difference between a website for e-commerce and one for reading a blog. This distinction significantly impacts the North Star metric focus. For example, if you are working on an e-commerce platform, the metric focus could be spending per month, while for a blog, you might focus on spending time on the page.

Now, let’s see an overview. In this case, we will focus on the Ramakien Game.

Game Overview Flow:

Currently, the Ramakien game only exists on the website.

• There are 3 ways to enter the game from the website.
• 2 Times story introduction and game onboarding in the game
• 10 levels of the game until getting certification.

Overall flow

Set up Hypotheses

To explore hypotheses, sometimes all I need to do is play around with the product. Questions or assumptions come to mind as I explore, and applying my findings from the previous chapter also works! The more I learn about the product, the better I can scope hypotheses.

Here are my hypotheses which arise from both playing around with the product and from my findings before.

#1 Which entrance is the most attractive to reach Ramakien?

Hypothesis: Based on students lacking an email to log in, the non-log-in entrance has a higher number of interactions than the log-in entrance.

#2 There are 2 times for the introduction and onboarding but are both of them still engaging?

Hypothesis: Completing the introduction story and onboarding flow from end to end will indicate how engaging the introduction is.

We will know if this is true when:

1. Users spend time not less than 3 seconds per page due to reading.
2. We don’t see a high number of clicks on ‘Skip’ compared to the introduction step.

Example of onboarding game

#3 Is the Ramakien game too challenging?

Based on current stats from Google Analytics, people spend 22 minutes, which is higher than the benchmark.

Hypothesis: Time spent on each stage and the number of interactions will indicate the difficulty of each stage.

We will know if this is true when:

• Users spend less time on earlier stages (1–6/10) than the later stages of the mini-game (6–10/10).
• There is high interaction on the main game until they click play result.
• There is a high number of users who reach the failed attempts page when compared with other stages.

#4 Even though the game is challenging, is it still engaging?

Hypothesis: Long-time spending and a high number of interactions, but users continue with the games. This will indicate how engaging the game is.

We will know if this is true when:

• Users spend a long time on each stage.
• There is a high number of interactions.

But they still click the ‘Next’ tab to move to the next level.

Selecting a Tracking Approach

There are several analytics tools available, including Adobe Analytics, Microsoft Power BI, and Heap Analytics. Based on what I’ve observed in many small organizations, such as startups or small businesses, they often have limited resources and may not have stringent requirements for tool capabilities. Typically, when starting small, they opt for Google Analytics due to its free accessibility.

Another reason for choosing Google Analytics is our teams may have already initiated the setup process.

As mentioned earlier in the project challenges, they have started tracking overall website metrics but have not delved into tracking specific interactions in the game.

When it comes to tracking at the interaction level, such as ‘Clicks,’ this requires more customized event tracking. Based on my current knowledge, there are 2 methods for custom event tracking:

1. Google Tag Manager:

With this approach, it’s possible to implement event tracking without coding. However, there is a learning curve involved in understanding how to use Google Tag Manager (Gtag) and how to fire tags. This method, in my view, may not be considered quite easy.

2. Pairing with Developers:

Developers can write a single line of code to specify what to track and the label’s name to display on Google Analytics. Afterward, they should place this code at the appropriate location. While this method doesn’t require a learning curve for designers, it does involve a dependency on developers, as effective communication and collaboration are essential to achieving the same goal.

Challenges in the Tracking Journey

Encountering a URL Challenge

Initially, I intended to use Gtag because I could proceed without dependencies, and currently, I am learning within a safe space. However, upon exploring how the current product operates, I discovered that the Ramakien game employs the same URL link on every page. This poses a challenge for using Gtag, as it requires distinct URL links to differentiate various ‘Next’ buttons on each page.

Story Data Analytics

After understanding the limitations, I needed to pivot the approach. I sought assistance from S.Songklod IToon to track the code. Since this is not my daily work but more of an exercise, we have limited time constraints to collaborate and catch up. To bridge communication gaps, I tried writing a detailed story to derive value from this situation.

The 3 Elements of Data Analytics story:

1. Hypothesis: Understand the value of why we are tracking and how we plan to analyze.
2. Correlation: Identify other factors that impact the analysis.
3. Naming Event: Create a systematic naming convention for all events.

Recap of My Learning in the 2nd Chapter:

Great hypotheses lead the direction of analysis:

I used to think that spending time formulating hypotheses and writing them properly was time-consuming. However, I found that this helps predict outcomes and determine where to focus the analysis.

Finally, event tracking is live!

Retrospective: Based on the time constraints during our catch-up, developers decided to use their own naming convention. As a result, I needed to spend some time catching up and understanding the meaning of each label. But this couldn’t stop us from moving forward :)

Next Chapter

In the next chapter, it will be the scenario for which we have already completed all event tracking. The next step involves connecting all these insights together on the Google Analytics dashboards which I will conclude as a Chapter 3 later after I play around. :)

Chapter 3:

Self-Learning Data Analytics as a UX Designer: #3 Let’s do the most fun part! Summary a thing.

If you want to review how I collected hypotheses previously, please check here.

Self-learning about data analytics as a UX designer #1

Big thanks to all my supporters

Thank you for your encouragement, guidance, and valuable suggestions: Pete Chemsripong, Pee Tankulrat

Thanks for the valuable opportunity on playground data sources and your collaboration effort: S.Songklod IToon, Tann Hiranyawech