What the delay means
A delay is a deliberate, fixed offset applied before market data is published. With a 15-minute delay, a quote shown now reflects the market as it stood 15 minutes earlier. The data itself is genuine and accurate for its timestamp; it is simply released later. This is distinct from latency, the small unavoidable transport time that every feed has, which is measured in milliseconds rather than minutes.
Why delayed data exists
Delayed data is generally cheaper to provide and subject to fewer distribution constraints than a live feed. As a result, it is the common default for free tiers, public websites, and read-only displays, where the cost and obligations of real-time delivery are not justified. Publishing data on a delay is a deliberate choice that lowers the barrier to making market information widely available.
Where delayed data is appropriate
Delayed data is suitable wherever a value that is a few minutes old would not change a decision. Long-horizon research, dashboards that refresh periodically, educational tools, and informational displays all work well on it. The test is straightforward: if acting on a 15-minute-old price would not cause a problem, delayed data is sufficient.
Delayed, real-time, and end-of-day
Delayed data sits between two other freshness levels. Real-time data arrives as the market moves, delayed data trails it by a fixed offset, and end-of-day data is a single snapshot per session. Anything that depends on the current price, such as trading, execution, or live alerts, requires real-time data rather than a delayed feed.
Real-time and historical on SiftingIO
SiftingIO is built for real-time delivery, streaming aggregated prices over WebSocket with sub-100ms median latency from primary regions, alongside historical and end-of-day data over REST. Where a delayed or static view is sufficient, the same data can simply be displayed without the live stream, under one schema across every market.