Post Event Analytics - CBA Glossary
📖 Glossary

Post Event Analytics

Post-event analytics measure how the broadcast performed: how many people watched, where they watched from geographically, how long they stayed, what content made them leave, where video quality suffered, which Interactive Features drove engagement. This data informs decisions about future events and helps justify the production investment.

Raw viewership numbers (concurrent viewers, total viewers) are table stakes but don't tell the real story. A webinar with 5,000 concurrent viewers where average view duration is 8 minutes (out of a 60-minute event) reveals that most people dropped off after 8 minutes. A webinar with 500 concurrent viewers and 45-minute average view duration shows higher engagement. The second event is more successful despite lower concurrent numbers.

Quality metrics matter. Analytics track buffering events (how many viewers experienced stalls), bitrate switches (when viewers' quality changed), and quality issues. If 30% of viewers experienced buffering during the first 10 minutes, your adaptive bitrate streaming ladder might have been misconfigured or your CDN was overloaded. These patterns inform technical improvements for future events.

Geographic distribution shows where viewers came from—especially important for CBA serving the UAE and Middle East region. Analytics might reveal that 40% of viewers joined from India, 35% from UAE, 15% from Europe, 10% from elsewhere. This informs CDN strategy—ensuring you have edge servers in high-traffic regions so viewers experience low latency.

For esports broadcasts and event streaming, engagement metrics matter. Which camera angles triggered the most chat activity? When did viewers drop off—during specific matches or during sponsorship announcements? Did interactive polls increase engagement or distract from broadcast? These insights drive improvements to interactive features and production aesthetics.

Revenue metrics matter for monetized streams—ad impressions, donation totals, subscription conversions. Even for non-monetized corporate events, analytics track cost per viewer, cost per engagement minute, and return on production investment. Did the 40-person production team generate ROI through lead generation or brand impact?

We integrate analytics from multiple sources: CDN analytics (bandwidth, geographic distribution), video platform analytics (YouTube, Twitch, proprietary platforms), interactive platform analytics (polls, chat activity), and custom analytics from our encoding and delivery systems. Synthesizing these into actionable insights requires experienced analysts.

FAQ
What should we measure from our broadcast events?
+
Start with: concurrent viewers, total viewers, average watch duration, geographic distribution, and buffering events. Layer in engagement metrics (chat activity, poll participation) if you're using Interactive Features. If monetized, add revenue metrics. Not every metric matters for every event—a corporate internal event prioritizes engagement; a public stream prioritizes geographic reach.
How do we know if our stream quality was good?
+
Through metrics. Target buffering rate below 2% (fewer than 1 in 50 viewers experienced a stall). Bitrate switches should be infrequent for stable internet connections; frequent switches indicate your bitrate ladder has gaps or your CDN was inconsistent. Viewer complaints on social media or direct feedback matter too, but objective metrics are more actionable.
Can we tell which camera angles were most watched?
+
Only if viewers make explicit choices (clicking between multiple concurrent streams) or if you use heatmap analysis that tracks where on-screen viewers focus. Standard analytics don't automatically show this. For esports, we sometimes use engagement spikes in chat to infer which camera angle was on air during high activity. This is correlation, not certainty.
How long after an event can we get analytics?
+
Real-time analytics start immediately (current viewer count, buffering events). Detailed analytics take 24-48 hours for CDNs and platforms to fully aggregate. Geographic data, demographic information, and engagement patterns take longer as data pipelines process. We typically deliver preliminary analytics within 24 hours and detailed reports within a week.

Need help with post event analytics?

Book a Discovery Call