How to Update Data in Your FlutterFlow Firestore Database

In FlutterFlow, select the widget that triggers the update (e.g., a Save button). Open its Action Flow. Add a Backend Call action → Update Document. Select the collection you want to update. For the document reference, either pass it as a Page Parameter (if navigating from a list) or fetch it via a Backend Query first. Under Fields to Update, click Add Field for each field you want to change. Set the field name and bind the value from a TextField or Page State variable. Fields you do NOT list here remain unchanged — this is a partial update, not a full document replace.

When multiple users can update the same numeric field (view counts, like counts, stock quantities), reading the value, adding 1, and writing it back creates race conditions. Two users reading simultaneously both see 5, both write 6 — one increment is lost. Fix this with a Custom Action using FieldValue.increment. In FlutterFlow, go to Custom Code → Custom Actions → Add Action. Import 'package:cloud_firestore/cloud_firestore.dart'. Write the function body: await FirebaseFirestore.instance.collection('posts').doc(postId).update({'viewCount': FieldValue.increment(1)}). Call this action from your Action Flow instead of the standard Update Document action for counter fields.

1// Custom Action: incrementField
2// Parameters: collectionPath (String), docId (String), fieldName (String), incrementBy (double)
3import 'package:cloud_firestore/cloud_firestore.dart';
4
5Future incrementField(
6  String collectionPath,
7  String docId,
8  String fieldName,
9  double incrementBy,
10) async {
11  await FirebaseFirestore.instance
12      .collection(collectionPath)
13      .doc(docId)
14      .update({fieldName: FieldValue.increment(incrementBy)});
15}

For fields that are arrays (tags, liked post IDs, category selections), avoid reading the array, modifying it in Dart, and writing the whole array back — this loses concurrent changes. Use two Custom Actions instead. arrayUnion adds an item only if it does not already exist (no duplicates). arrayRemove removes all instances of the specified item. Create a Custom Action named toggleFavorite that takes a productId String. Inside: if adding, call .update({'favoriteIds': FieldValue.arrayUnion([productId])}). If removing, call .update({'favoriteIds': FieldValue.arrayRemove([productId])}). Pass a boolean parameter isFavoriting to choose which path to take.

1// Custom Action: toggleFavorite
2// Parameters: userId (String), productId (String), isFavoriting (bool)
3import 'package:cloud_firestore/cloud_firestore.dart';
4
5Future toggleFavorite(
6  String userId,
7  String productId,
8  bool isFavoriting,
9) async {
10  final userRef = FirebaseFirestore.instance
11      .collection('users')
12      .doc(userId);
13  if (isFavoriting) {
14    await userRef.update({
15      'favoriteIds': FieldValue.arrayUnion([productId]),
16    });
17  } else {
18    await userRef.update({
19      'favoriteIds': FieldValue.arrayRemove([productId]),
20    });
21  }
22}

If you need to update several documents together — for example, marking all messages in a conversation as read, or updating a product's price across a cart and an order — use a WriteBatch so either all updates succeed or none do. Create a Custom Action named batchUpdateReadStatus. Parameters: conversationId (String), userId (String). Inside: get a WriteBatch via FirebaseFirestore.instance.batch(). Query unread messages. For each, call batch.update(docRef, {'readBy': FieldValue.arrayUnion([userId])}). After the loop, await batch.commit(). Firestore WriteBatch supports up to 500 operations per commit. If you need more, split into multiple batches.

1// Custom Action: batchMarkRead
2// Parameters: conversationId (String), userId (String)
3import 'package:cloud_firestore/cloud_firestore.dart';
4
5Future batchMarkRead(
6  String conversationId,
7  String userId,
8) async {
9  final messagesRef = FirebaseFirestore.instance
10      .collection('conversations')
11      .doc(conversationId)
12      .collection('messages');
13  final unread = await messagesRef
14      .where('readBy', whereNotIn: [[userId]])
15      .limit(500)
16      .get();
17  if (unread.docs.isEmpty) return;
18  final batch = FirebaseFirestore.instance.batch();
19  for (final doc in unread.docs) {
20    batch.update(doc.reference, {
21      'readBy': FieldValue.arrayUnion([userId]),
22    });
23  }
24  await batch.commit();
25}

When your update logic depends on the current value of a field — for example, booking a seat only if the available count is still above zero — you must use a transaction. A transaction reads a document and then updates it in an atomic sequence. If another write modifies the document between your read and write, Firestore automatically retries the transaction. Create a Custom Action named bookSeat. Inside: FirebaseFirestore.instance.runTransaction((transaction) async { final snap = await transaction.get(eventRef); final available = snap.data()!['availableSeats'] as int; if (available <= 0) throw Exception('No seats available'); transaction.update(eventRef, {'availableSeats': FieldValue.increment(-1)}); transaction.set(bookingRef, {'userId': userId, 'eventId': eventId, 'bookedAt': FieldValue.serverTimestamp()}); }). If you need expert help structuring complex transaction logic across multiple collections, RapidDev has implemented this pattern across hundreds of FlutterFlow apps.