Enabling Google Analytics

Enabling Google Analytics — GoTradie Applications

What is this article about?

In this article, I’m going to give a simple explanation of how I applied analytics techniques to the GoTradie applications. Also, problems that I have faced and what are the plants that I’m planning to do in the improvements in this area for GoTradie applications.

Why did we apply data analytics systems to our applications?

Collecting data from the users is essential thing when comes to maintaining and updating websites or mobile apps according to user interactions and user preferences. In GoTradie that’s the main reason for applying data analytic techniques for our applications.

What did we use?

There are many tools and methods when comes to applying data analytics to an application. When considering using analytics we considered factors like ease of implementation, accuracy, scalability, and customization ability. When considering that we used google analytics to get the data that we needed because we saw that it have all those features.

What are the metrics that we wanted and how we organized the GA structure?

Here we needed to gather analytical data about users’ behavior when using GoTradie’s web and mobile applications. Also need to mention that we had three types of signup ways in our web application.

1. Normal signing — This is done by entering user data into the signup form.
2. By claiming businesses — In here user needs to have an invitation link by a user who has created a business(account) in the GoTradie app or web application.
3. The Prometheus admin portal — In here user gets the claim business link (invitation link) from the admin portal named Prometheus.

So we had to implement analytics support to each signup method to get more accurate data that wanted. So these are the details we needed by implementing the analytics to the mobile and web applications.

• Users ( Each according to all the platforms, by mobile and by web app)
  - How many joined in the last 7 days, 14 days, 1 month, etc
  - Total number of users
• Daily Active Users(DAU) and Monthly Active Users(MAU). ( Each according to all the platforms, by mobile and by web app)
  - All the users by phone number of the user.
  - Drilled down by Business.
  - Drilled down by City.
  - Drilled down by the trade.
  - Drilled down by user type.
  - Drilled down by user type. (Owner, Admin, Member, etc.)
• Posted clips count.
  - Drilled down by users.
  - Drilled down by businesses.
• Sent messages count.
  - Drilled down by users.
  - Drilled down by businesses.
  - Drilled down by day.
• To-Do lists creation count.
  - Drilled down by users.
  - Drilled down by businesses.
  - Drilled down by day.
• Tasks lists creation count.
  - Drilled down by users.
  - Drilled down by businesses.

To get these pieces of information we implemented some matrics. These are as follows.

• Total users.
• 1-day active users.
• 7-day active users.
• 28-day active users.
• Event count per user.
• Event count.

Also not only that we used dimensions to sort out and segment our data. Those are…

• phone_number( This is the user’s phone number. So we used this to identify the user)
• org_id( In long form organization ID. Which means the business ID)
• Date

Why did we use Google Looker Studio?

So after using google analytics we needed to show our dashboards in a clear way to the users who views our dashboards. For that, we had two options.

1. Google Analytics exploration section.
2. Using Looker Studio (Google Data Studio in the past).

When considering those two options we saw that using there are some advantages of using Google Looker Studio over the Google Analytics exploration section. Able to create user-convenient dashboards, more control in the dashboards, and customization ability among them. That’s the reason behind choosing Google Looker Studio over the Google Analytics exploration section to show metrics that we got from using Google Analytics.

What are the different dashboards that we have?

In the Google Looker Studio, we organized our dashboards in a convenient way. First rather than creating all the dashboards on the same page in the google looker studio we created five pages for that. Dashboards and pages which we have created are mentioned below,

1. Overview.
   - All users with phone number.
   - All users with company ID.
2. Users’ login and signing behavior.
   - Total users, Active users in mobile app and web app.
   - Total users, Active users in the mobile app.
   - Total users, Active users in the web app.
   - Active users (last 7 days).
   - Active businesses (last 7 days).
3. User signups.
   - Users signup via all platforms.
   - Users signup via the mobile app.
   - Users signup via the web app (Normal + Claim).
   - Users signup via the web app (Claim).
   - Cumulative users signup via all platforms.
   - Cumulative users signup via the mobile app.
   - Cumulative users signup via the web app (Normal + Claim).
   - Cumulative users signup via the web app (Claim).
4. Messages and clips.
   - Messages count by each user.
   - New clips count by each user.
   - Messages count by each day.
   - New clips count by each day.
5. To-Do lists and Tasks.
   - To-Do lists creation count by each user.
   - Tasks creation count by each user.
   - To-Do lists creation count by each day.
   - Tasks creation count by each day.

Unique user challange?

Challenges always happen when doing something. When implementing google analytics there were some challenges that we had to face. When considering the accuracy of the analytics, we had to face a very huge problem. When getting the total users, we had an issue getting the total users to count as more than the expected count (actual count). There were several reasons behind that issue.

• Users log in from different devices.
• Users log in from different browsers.
• Users clear cache after using the application.

So this happens because in google analytics, google stores a cookie in the cache to detect the user. So when a user comes to the site on different devices or from different browsers the cookie data get changed. Also, the same thing happens when clearing the cache of the browser. Because of that, we needed to implement a way to detect unique users who are coming to the website. For that, we found a way to detect unique users by sending a unique user ID after users signup or log in. When a user gets signup for the first time when a user gets logged in a userID will be sent to google analytics. So from that google analytics is able to detect the unique users in our application when a user changes their devices or browsers and even when they clear their cookies.

What’s next?

So as the next step is that we hope to get more accurate and deep analytical data by using deep data analytical methods. For that as a deep data analytical method, we are planning to use Twilio Segment with Amplitude analytics. So by using Twilio Segment, we can collect user interaction data from multiple platforms, and with the support of Amplitude analytics, we can get the required analytical data.