Blog News & Product updatesWeld Product Updates – March 2026

News & Product updatesApril 01, 2026 • 5 min read

Weld Product Updates – March 2026

This month, Weld launched Klarna and PayPal connectors, a new Usage Dashboard with AI-driven forecasting, dbt-compatible Model Import/Export, and major improvements across destinations and data sources.

by Jonas Thordal

Weld Recap March 2026: Klarna & PayPal Connectors, Usage Dashboard, dbt Model Import & More

Spring is here, and we're bringing a massive wave of updates to Weld! This month, we're thrilled to introduce new connectors for Klarna and PayPal, a brand-new Usage Dashboard, dbt-compatible Model Import/Export, and significant reliability boosts across the platform.

Here's everything new and improved in Weld this month.

New Features

Klarna Connector

You can now seamlessly extract data from Klarna. We've added full sync support for transaction, payout, and payout_summary streams.

Explore Klarna documentation →

3 total changes

Table/column	Change type	Notes
`transaction`	New table
`payout`	New table
`payout_summary`	New table

PayPal Connector

Connect your PayPal account to Weld to automatically sync balances, invoices, transactions, billing_plans, subscriptions, and disputes.

Explore PayPal documentation →

6 total changes

Table/column	Change type	Notes
`balances`	New table
`invoices`	New table
`transactions`	New table
`billing_plans`	New table
`subscriptions`	New table
`disputes`	New table

Model Import & Export (with dbt Support)

Managing and sharing your SQL models is easier than ever. You can now import and export your models as zip files directly through the UI or API.

Exporting supports two formats:

Weld-Native — a zip containing .sql files and a manifest.json with dependencies, materializations, and scheduling metadata.
dbt-Compatible — a complete dbt project with dbt_project.yml, sources.yml, and Jinja {{ ref() }} / {{ source() }} syntax ready for dbt Cloud or dbt Core.

Importing works in reverse — upload a .zip and Weld auto-detects the format. dbt projects are automatically converted into Weld-native dependencies, with a dry-run preview before anything is committed. Folder structures are preserved as model paths, and dependencies are resolved against your existing workspace.

This makes it straightforward to:

Migrate models between Weld workspaces
Back up your transformation layer as versioned zip files
Transition to dbt by exporting a ready-to-run dbt project
Onboard from dbt by importing an existing dbt project into Weld

Explore Model Import & Export documentation →

Usage Dashboard & Forecasting

We've launched a new Usage Dashboard right in the Weld app. You can now easily track your Monthly Active Rows (MAR) and synced rows, view historical usage trends, and see AI-driven forecasts of your future data consumption.

Connector Updates

Google Search Console & Reddit Ads

We've unlocked custom reporting capabilities for both platforms. You can now configure your own dimensions, aggregations, and breakdowns directly in your integration settings.

Amazon Selling Partner & Amazon Ads

New API: Upgraded to the newer searchOrders API for more reliable order syncing.
Unified Reports: Added hourly campaign reports, associated orders, and new buyer details (email, name, PO number) to your syncs.

4 total changes

Table/column	Change type	New name	Notes
`order`
`column`	New Column	`buyer_email`
`column`	New Column	`buyer_name`
`column`	New Column	`purchase_order_number`
`hourly_campaign_report`	New table		Unified hourly campaign reports

Shopify

Faster Syncs: Optimized incremental syncs for faster product updates.
New Fields: Added lineItem details to order agreements and featured_media_id to products.

2 total changes

Table/column	Change type	New name	Notes
`order_agreement`
`column`	New Column	`lineItem`	Line item details on order agreements
`product`
`column`	New Column	`featured_media_id`

HubSpot

New Stream: Added account_details stream.
Marketing Email: Expanded Marketing Email content.
Custom Objects: Introduced support for generic v4 Custom Objects association tables.

1 total change

Table/column	Change type	Notes
`account_details`	New table

Facebook Ads & Microsoft Ads

Multi-Account Backfill: Added multi-account backfill support.
Resilience: Improved API resilience to prevent transient errors from failing syncs.
Microsoft Ads Fix: Fixed report parsing to natively exclude headers and footers.

TikTok Shop

New Stream: Introduced STATEMENT_TRANSACTION_REPORT to fix an issue where legacy finance tables were showing up empty.

1 total change

Table/column	Change type	Notes
`STATEMENT_TRANSACTION_REPORT`	New table	Replaces legacy finance tables

Google Ads

API Upgrade: Upgraded the integration to the v23 API.
New Stream: Added SEARCH_TERM_STATS.
Quality Scores: Added quality_info (including Quality Scores) to ad group criteria.

2 total changes

Table/column	Change type	New name	Notes
`SEARCH_TERM_STATS`	New table
`ad_group_criteria`
`column`	New Column	`quality_info`	Includes Quality Scores

More Stream Additions

Zendesk: Added ticket_side_conversation and ticket_side_conversation_event streams.
GoHighLevel: Added conversation and message streams.
AppLovin: Added new daily and hourly advertiser_campaign_report streams.
Xero: Added a toggle for advanced plan features and more granular permission scopes.

6 total changes

Table/column	Change type	Notes
`ticket_side_conversation`	New table	Zendesk
`ticket_side_conversation_event`	New table	Zendesk
`conversation`	New table	GoHighLevel
`message`	New table	GoHighLevel
`advertiser_campaign_report_daily`	New table	AppLovin
`advertiser_campaign_report_hourly`	New table	AppLovin

Platform & Destination Improvements

Smarter Reverse ETL

Reverse ETL syncs are now much more resilient. Known, row-level destination errors (like a single missing user in an external app) will be tolerated and logged, rather than triggering a complete sync failure. We also dramatically sped up Excel table replacements to eliminate timeouts.

BigQuery Zero-Downtime Updates

We introduced an atomic swap mechanism for BigQuery. When table partitioning changes, Weld now creates and swaps the tables seamlessly behind the scenes, eliminating downtime and race conditions. We also added automatic cleanup for temporary tables.

Snowflake Reliability

Fixed a timeout issue that could cause a dead loop when syncing exceptionally large files to Snowflake, and added intelligent upload retries to make destination syncs bulletproof.

Postgres Optimization

Prevented timeouts during massive MERGE operations and optimized full syncs to skip temporary primary key overhead, making heavy Postgres loads much faster.

CDC for Free Trials

Change Data Capture (CDC) replication is now enabled for free trial accounts. We also made under-the-hood engine improvements to prevent timestamp collisions and improve data ordering.

Retired Stream Tags

Data streams that have been deprecated by their platform are now visually tagged as "Retired" in the Weld UI, so you know exactly when it's time to migrate to a newer table version.

Beyond these highlights, we also shipped numerous bug fixes and reliability improvements across pagination, timestamp parsing, and connector authentication. For the full technical changelog, check out the March 2026 Release Notes.

That's it for March. If you have ideas, questions, or want to see something in Weld — let us know.