Concurrency is the number of viewers watching at the same moment — not total accounts or daily users, but how many streams are live right now. It is the single most important number for capacity and cost, because delivery load is driven by simultaneous demand, not by library size or registered users. A service with ten million subscribers might peak at a few hundred thousand concurrent streams on an ordinary night and many times that for a marquee live event.
Concurrency drives almost every scaling decision: CDN egress (the dominant cost), origin and origin-shield capacity, license-server throughput, and the headroom needed to survive the synchronized spike when a premiere or kickoff starts and everyone presses play within the same minute.
Because VOD demand is spread across time while live demand collapses into a single instant, the same audience can produce wildly different peak concurrency depending on content type — which is why live events, not catalog size, usually set the high-water mark a platform must engineer for.
The handbook · 7 volumes
Fora Soft Video Engineering Handbook
Concurrency is what your origin and CDN feel at peak, and a big live event can multiply it in minutes. We wrote seven volumes on the video stack; for concurrency planning start with Volume 3 on CDNs, caching, and the economics of delivery, then Volume 2 on ingest and live packaging, with encoding in Volume 1. Written by Nikolay Sapunov, CEO at Fora Soft.
Preparing your platform for a concurrency spike during a big live event?
We will model your peak load and send a scaling plan for origin, CDN, and players with a cost per event. 250+ video products shipped since 2005.
