For some realtime applications, the latest state matters more than the full stream of updates. If you’re tracking stock prices, monitoring live sports scores, or displaying GPS locations, you don’t need to process and transmit every intermediate update—just the most recent one.
To help you minimize bandwidth usage, processing costs, and system load, we’ve introduced conflated subscriptions to Ably Pub/Sub.
With conflated subscriptions, subscribers receive only the latest message within a defined conflation window, as outdated updates are automatically discarded when a newer message replaces them.
The challenge: High-frequency updates can lead to excessive message processing
Many realtime systems generate high-frequency updates—especially in data streams where values change rapidly. However, not all updates are equally important.
Take a stock price dashboard as an example. If the stock price fluctuates 100 times per second, but the display only updates every 100 milliseconds, there’s no reason to process and transmit every change. Instead, only the most recent value in each update period needs to be sent to clients.
Introducing conflated subscriptions
Conflated subscriptions operate at the channel level, automatically reducing message volume for clients that only need the latest data.
- Developers define a conflation window (e.g., 100ms, 200ms).
- Within that window, only the most recent update is delivered—earlier updates are discarded.
- Message delivery remains seamless, ensuring clients receive timely updates without unnecessary data.
In our stock example, without conflation, every individual update is sent, even if it’s outdated by the time it reaches the user. Subscribers process redundant messages, increasing CPU load, and bandwidth costs scale unnecessarily with the frequency of your updates.
But with conflation, only the latest update in each conflation period is sent, subscribers process fewer messages, and bandwidth costs decrease, improving efficiency at scale.
Benefits at a glance
- Reduces message volume: Subscribers receive only the latest update per conflation period, reducing redundant processing.
- Optimizes bandwidth usage: Less data transmission means lower network costs and improved efficiency.
- Decreases computational load: Subscribers handle fewer messages, leading to faster UI rendering and improved system performance.
- Simplifies development: No need for you to manually filter redundant updates—Ably handles it at the infrastructure level.
Conflated subscriptions use cases
Conflation is ideal for scenarios where realtime updates are frequent, but only the latest state matters, like in sports scoreboards for fan engagement use cases, where you can deliver the current score at controlled intervals without overwhelming the frontend with every micro-event. This also works for things like stock price dashboards (as we saw), live location tracking, and analytics dashboards, which all operate on the principle of requiring periodic refreshes while not necessarily needing every single update.
Get started with conflated subscriptions
Conflated subscriptions are available now in Ably Pub/Sub and can be enabled at the channel level with a simple configuration. Head to our docs to quickly walk through the process. If you have any questions, please get in touch - we’ll be happy to help.
