Tracking the Performance of TEDx videos

Ever wondered how to gain insights into the performance of TEDx talks on a YouTube channel you don’t own? Join me on a journey as we explore the evolution of a TEDx video tracker that has now migrated to the cloud, making it accessible to organizers worldwide.

The TEDx Case: A Global Movement in Focus

TEDx, a dynamic global movement dedicated to Technology, Entertainment, Design, and independently organized events, stands as a beacon for spreading innovative ideas and fostering meaningful discussions. These events showcase compelling talks by inspiring speakers, covering a diverse array of topics. The talks, uploaded to a central YouTube channel, serve as a valuable resource for viewers globally.

As TEDx organizers, we are naturally keen on monitoring the performance of our videos. Whether to gauge audience interest, understand YouTube trends, or celebrate milestones with our speakers, tracking video analytics is crucial. However, without direct access to channel analytics, organizers often resort to manual tracking of their most relevant videos.

The Birth of TEDx YouTube Tracker

Five years ago, I introduced the TEDx YouTubeTracker, a Python script that logged data for a specific TEDx group based on its name. Running almost daily on our server for five years now, this script provided valuable insights for TEDxTUM. For instance, it revealed the surge in views for Ralph Ammer’s talk, ‘How drawing helps you think,’ during the initial months of the 2020 pandemic. It also highlighted interesting patterns, such as videos with fewer views initially experiencing sudden increases, often triggered by media reports about our speakers.

Over time, I enhanced the functionality, incorporating features to plot data and even display a web dashboard. While the tracker generated interest, its adoption remained limited due — i assume mostly to the high technical entry barrier. To date, TEDxTUM stands out as the only TEDx actively tracking their video statistics on a daily basis using TEDx-YTT.

Transformation to the Cloud

In an effort to democratize access and make the TEDx video tracker more accessible, I migrated it to the cloud. As a welcome weekend project, I migrated the code to run on the Google Cloud Platform (GCP). The result? The TEDx YouTube Tracker (TEDxYTT) is now much simpler to set up, even without your own server.

The GCP iteration of the YouTube tracker empowers users to craft personalized reports and dashboards through Looker Studio — a tool making the design of data dashboards as simple as building a powerpoint slide. Explore an illustrative report featuring TEDxTUM data here, or browse the provided screenshots below.

Page 1 of the report, displaying the most current data in table format, including links to the videos on YouTube.

Page 2 of the report, displaying video performance over time. The videos to be displayed can be selected by choosing speaker, title or year published and the time span to be displayed can also be selected.

Page 3 of the report, displaying the average views, comments and likes over time. Other statistical measures (e.g. 25% percentile) can also be selected.

The code, along with a comprehensive setup guide, can be found in this GitHub Repository. All you need is basically a Google account and up to 1 US$ per month to spare (estimated cost based on my experience). If you’re a TEDx organizer struggling with the setup, don’t hesitate to reach out. Let’s empower TEDx communities worldwide with data-driven insights and analytics, fostering a deeper understanding of the impact of our talks on a global stage.

Technical Setup

For those who want to know a little more, here is the overarching technical setup on Google Cloud Platform (GCP) for the TEDx YouTube Tracker:

1. Cloud Scheduler and Pub/Sub Integration:

Cloud Scheduler orchestrates the process by publishing a message (‘update’ or ‘search’) to a Pub/Sub topic.

2. Cloud Function Trigger:

A Cloud Function, triggered by EventArc, responds to the message from Cloud Scheduler. It dynamically initiates either an update of existing video data or a search for new videos, adapting to the workload specified in the message.

3. Securing YouTube API Key:

The YouTube API v2 key is securely passed to the script using SecretManager, eliminating the need to store the API key in clear text.

4. Data Storage:

Data is persistently stored in BigQuery tables for seamless accessibility. Additionally, data is archived as *.csv files on Cloud Storage, ensuring redundancy and flexibility in data retrieval.

5. Integration with Looker Studio:

Looker Studio seamlessly connects with BigQuery, offering an intuitive interface for visual data exploration and analysis.

For a comprehensive guide on implementing video tracking for a YouTube channel that isn’t your own, delve into the technical intricacies detailed in the TEDx YouTube Tracker (TEDxYTT) GitHub Repository.

If you’re keen on venturing into the realm of YouTube Data API for your own analyses, this article serves as an excellent starting point.

Start Tracking Your Video Performance

The TEDx YouTube Tracker, born from a desire to automate some tedious video analytics tracking for TEDxTUM, has transformed into a cloud-based ally accessible to all TEDx organizers. I hope migrating the TEDx YouTube Tracker, making it more accessible for organizers worldwide, will also drive its adoption and provide insightful data.

Happy tracking! 🚀