Video pricing quoted per gigabyte hides what streaming businesses actually buy: audience minutes. Translating between the two is simple arithmetic that surprisingly few quotes bother to show.
The translation
A stream’s bitrate ladder decides how many gigabytes a viewer-hour consumes: higher quality, more bytes. Multiply average delivered bitrate by watch time and the per-GB rate becomes cost per viewer-hour, the unit your revenue already thinks in. Any provider comparison should happen in that unit. The unit mismatch is not accidental: gigabytes are what providers meter, viewer-hours are what businesses sell, and whoever does the translation controls the comparison.
Where video pricing hides costs
Regional rate spreads matter more for video because audiences concentrate. Request fees are negligible for long segments and meaningful for short-segment low-latency modes. And peak commitments for live events price the audience you hope for, which deserves honest forecasting rather than optimism. Short-segment low-latency modes deserve their own line in any model: chasing broadcast-grade latency multiplies request counts, and on request-priced networks that choice has a price tag worth knowing before the product decision, not after.
Codec strategy belongs in the same spreadsheet, because compression efficiency is delivery money. Modern codecs cut bitrates substantially at equivalent quality versus legacy encoding, and every saved megabit multiplies across the entire audience for the life of the catalog. The trade is encode cost and device-compatibility engineering against perpetual delivery savings, and at audience scale the delivery side wins decisively. A streaming business negotiating hard on per-GB rates while shipping legacy-encoded bitrates is optimizing the small factor: the ladder itself is usually the bigger lever, and it is entirely under your control.
Buying delivery for a media business
Committed tiers on your base load, burst arrangements for events, multi-CDN where the audience or the stakes demand it, and everything modeled per viewer-hour so finance and engineering argue about the same number. This is the shape of every well-bought streaming contract we have seen. Peak-versus-base separation is the structural insight: paying committed rates for your predictable floor and negotiated burst for your spikes beats sizing one commitment for a hope.
In practice
Build the one spreadsheet that matters: your bitrate ladder, the audience share on each rung, average watch time, and each provider’s rate structure. Out comes cost per viewer-hour per provider, the number your unit economics already speak. Every streaming procurement we run starts by building exactly this sheet, and it ends most pricing debates before they start, because both sides finally argue about the same number.
Our streaming assessment returns your cost per viewer-hour across providers, from your actual bitrate ladder.