Skip to main content

​
Overview

Snag maintains an immutable loyalty ledger and accurate per-currency balances for each user. We do not store or expose the total points a user earned from a single loyalty rule (“rule-level points”).
We don’t provide any endpoint to fetch “points by rule” for one user or a list of users. If you need rule-level totals, you should track and store them in your own system.
What we do provide: a complete transaction stream and balance updates. You can reliably derive per-rule totals by consuming transaction events and persisting aggregates keyed by (userId, loyaltyRuleId) on your side.
At a high level, you should:
  • Ingest new transactions in real time via subscriptions → webhooks
  • Periodically reconcile using the transactions listing endpoint
  • Maintain idempotent, append-only processing in your datastore
1

Receive new transactions via webhooks

Enable a subscription for LoyaltyTransactionEntry and deliver events to your webhook endpoint. See Subscriptions and Webhooks.
LoyaltyTransactionEntry may not always include a loyaltyRuleId (e.g., manual adjustments). Skip entries without a rule when calculating rule-level totals.
2

Reconcile periodically

Run a periodic job to fetch recent entries and backfill anything missed by your webhook consumer. See Get Loyalty Transaction Entries.
3

Store minimal aggregates on your side

  • Keep a lightweight aggregate keyed by (userId, loyaltyRuleId)
  • Apply credits and debits accordingly
  • Deduplicate using entry.id or idempotencyKey

​
FAQ

No. Snag’s source of truth is the ledger and account balances. Aggregations by rule are intentionally left to partner systems so you can shape them to your business logic and retention needs.
Use entry.id or idempotencyKey to deduplicate. Keep a simple record of processed entries and make your updates idempotent.
Entries include a direction of credit or debit. Apply debits as negative adjustments to your stored totals.
This is why the periodic reconciliation step exists. Use startingAfter with the last processed entry.id to fetch any missed entries and catch up exactly once.
You now have a reliable pattern to calculate, store, and query rule-level point totals using Snag’s transaction stream.