Sessions & Events

Understanding sessions and events is fundamental to using Amply SDK effectively. This guide covers how they work and best practices for implementation.

Sessions

A session represents a period of user activity within your app. The SDK manages sessions automatically, but understanding their lifecycle helps you build better analytics.

Session Lifecycle

App Launch → Session Start → User Activity → Inactivity → Session End
                                    ↓
                              New Activity → Session Resume or New Session

Session Properties

Each session tracks:

Property	Description
counter	Total number of sessions for this user
duration	Time elapsed since session start
eventCount	Number of events in current session

Accessing Session Data

Kotlin

// Get session dataset snapshot
val sessionData = amply.getDataSetSnapshot(DataSetType.Session)
// Returns: { "counter": 5 }

Swift

Events

Events capture discrete user actions and system occurrences. They form the basis for analytics and campaign triggering.

Event Types

Type	Description	Example
Custom	User-triggered actions	button_clicked, purchase_completed
System	SDK-generated events	SDK_INITIALIZED, SESSION_START

Tracking Custom Events

Kotlin

// Simple event
amply.track("screen_viewed")

// Event with properties
amply.track("item_purchased", mapOf(
    "item_id" to "SKU123",
    "price" to 29.99,
    "currency" to "USD"
))

Swift

System Events

The SDK automatically tracks these system events:

Event	When Fired
SDK_INITIALIZED	SDK initialization complete
SESSION_START	New session begins
SESSION_END	Session ends
CAMPAIGNS_LOADED	Campaigns fetched from server

Listening to System Events

Kotlin

// Android
amply.setSystemEventsListener { event ->
    Log.d("Amply", "System event: ${event.name}")
}

Swift

Event Properties

Properties add context to events. They can be used in campaign targeting and analytics.

Property Types

Events support these property value types:

• String

• Int / Long

• Double / Float

• Boolean

Best Practices

1

Use consistent naming

Stick to a naming convention (e.g., snake_case).

2

Keep properties flat

Avoid nested objects.

3

Be specific

purchase_completed is better than action.

4

Include context

Add relevant IDs, categories, values.

Example Event Schema

Kotlin

// Good event design
amply.track("product_viewed", mapOf(
    "product_id" to "12345",
    "product_name" to "Premium Widget",
    "category" to "widgets",
    "price" to 49.99,
    "currency" to "USD",
    "source" to "search_results"
))

Event Counters

The SDK maintains counters for events, useful for campaign triggering:

Counter Type	Scope	Use Case
Global	All time	"Show after 10th purchase ever"
Session	Current session	"Show after 3rd view this session"

Accessing Event Counters

Kotlin

// Get triggered event data with counters
val eventData = amply.getDataSetSnapshot(
    DataSetType.TriggeredEvent("purchase_completed")
)
// Returns: { "name": "purchase_completed", "counter": 5, "sessionCounter": 2, ... }

Retrieving Recent Events

Kotlin

// Get last 30 events
val recentEvents = amply.getRecentEvents(limit = 30)

recentEvents.forEach { event ->
    println("${event.name} at ${event.timestamp}")
}

Swift