What is SDK event latency?
Used in a sentence: The cart nudge shows up after users have already left checkout, so we need to look at our SDK event latency before changing the copy.
How it works
When a user taps, scrolls or opens a screen, the app logs an event. Latency is the time it takes for that event to reach whatever decides on a response, get evaluated against campaign rules, and produce something on screen.
The biggest factor is where the decision happens. If the SDK holds campaign rules on the device, it can match the event and render the message almost instantly. If the event must travel to a server, join a queue and wait for a segment refresh, the delay can stretch from seconds to hours. Content loading adds more delay, which is why good SDKs fetch campaign assets in advance.
Example
A quick commerce app wants to show a free delivery offer when a user removes an item and the cart drops below the free delivery threshold. With a batch-processed setup, the offer arrives minutes later as the user is closing the app. With on-device evaluation, the bottom sheet appears the instant the item is removed, while the user is still looking at the cart.
Real-time triggers vs batch-processed segments
| Real-time triggers | Batch-processed segments | |
|---|---|---|
| Decision made | On the device, at the event | On a server, on a schedule |
| Delay | Imperceptible | Minutes to hours |
| Best for | In-session messages | Daily or weekly targeting |
Batch segments are fine for deciding who is eligible. Real-time triggers decide when to show it.
Best practices
- Evaluate trigger rules on the device wherever possible.
- Preload campaign content so rendering does not wait on the network.
- Measure the gap between event time and display time, not just delivery.
- Keep the SDK small and test its effect on app start time.
How to do this with Digia Engage
Digia Engage triggers campaigns on real user events in under 100ms, so messages appear while the user is still on the screen. The SDK is under 2MB and integrates in about 20 minutes on iOS, Android, Flutter and React Native. It reuses events already tracked in CleverTap, MoEngage or WebEngage, so teams do not need a separate tracking plan.
Frequently asked questions
- What is a good event latency for in-app messages?
- Low enough that the user cannot notice a gap between their action and the message, which means well under a second. In practice, the message should appear before the user can scroll, tap away or change screens. Anything that relies on a round trip to a batch system will usually be too slow.
- Does an engagement SDK slow down my app?
- A well-built one should not. Look at the SDK's size, whether it evaluates rules on the device, and whether it loads campaign content in the background rather than on the main thread. Test app start time and screen load with and without the SDK before rolling it out widely.
Related terms
Read more
Part of the In-app engagement glossary.