Latency vs. Interactivity: The Tradeoff Every Live Producer Has to Pick
Sub-second glass delay buys real-time audience response — and costs you CDN scale. Twenty seconds of delay buys reach — and kills the chat. The decision framework.
Every live production makes an unspoken choice: how far behind reality the audience sits. That delay — “glass-to-glass” latency — is not a technical detail. It’s a creative and commercial decision disguised as one, and the tradeoff is genuine.
Lower latency costs scale. Higher latency costs interaction. There is no free option.
The Latency Ladder, Honest Version
| Glass delay | What it enables | What it costs |
|---|---|---|
| < 0.5s (WebRTC) | Real-time chat response, live auctions, synced reaction | Expensive egress, complex scaling |
| 1–3s (SRT→WebRTC hybrid) | Conversational back-and-forth | Still premium delivery pricing |
| 5–10s (LL-HLS) | “Live” feel, poll participation | Cheap-ish, works at scale |
| 20–45s (standard HLS) | Broadcast-grade reliability | Chat feels like a tape delay |
Pick by Interaction Type, Not Preference
- Pure one-way broadcast (concert, keynote): latency doesn’t matter. Run standard HLS, take the scale.
- Audience-influenced live (Q&A, live shopping, gaming): the chat needs to feel like a conversation — sub-5s or the audience waits through your answer while you read it.
- Interactive / transactional live (auctions, live polls with stakes): WebRTC territory. Every extra second of delay is a decision window your audience can’t control.
“The right latency isn’t the lowest one you can afford. It’s the highest one that doesn’t break the interaction.”
Decision matrices for every format in the Latency vs. Interactivity Decision Guide.
Latency is a dial, not a defect. Turn it deliberately.