Campaign-to-Pipeline Attribution
Sync Google Ads spend and HubSpot campaign engagement into BigQuery on a schedule so marketing and sales report from the same pipeline numbers.
The problem
Marketing reports from the ad platforms, sales reports from the CRM, and the numbers rarely agree. Analysts export spend and engagement by hand into spreadsheets or the warehouse, so attribution is always days behind. Nobody trusts the pipeline number in the Monday meeting.
How it runs on Neblex
A cron schedule pulls spend and performance from Google Ads at whatever cadence you choose. Poll-based change detection watches HubSpot campaigns and contacts, catching inserts, updates, and deletes through cursors and content hashing, so removed or renamed campaigns do not linger in your reporting.
Transforms and field mapping shape both sources into a common model, and chunked batch processing with checkpoints loads the data into Google BigQuery. Because every database connector supports arbitrary SQL, the flow runs the join and attribution queries in BigQuery as a step, then writes summarized attribution back to HubSpot.
If a load fails partway, the run resumes from its last checkpoint instead of starting over, and a failed run can be replayed from the exact step that broke. Bulk replay lets you rerun a whole window of loads after a mapping change.
Step by step
Pull ad spend
A cron schedule fetches Google Ads spend and performance for the reporting window.
Detect CRM changes
Poll-based change detection picks up new, updated, and deleted HubSpot campaigns and engagements.
Transform to one model
Transforms and field mapping align both sources on shared campaign and contact keys.
Load into BigQuery
Chunked batches with checkpoints write the data, so partial failures resume cleanly.
Compute attribution in SQL
An arbitrary SQL step runs the attribution queries directly in BigQuery.
Write back to HubSpot
Summarized attribution lands on campaigns so marketing and sales read the same numbers.
Platform capabilities used
- Cron schedules
- Poll-based change detection
- Chunked batch processing with checkpoints
- Arbitrary SQL on database connectors
- Transforms and field mapping
- Bulk replay
Common questions
How are deleted or renamed campaigns handled?
Poll-based change detection uses cursors and content hashing, so it sees deletes and updates as well as inserts. The flow can mark removed campaigns in BigQuery instead of leaving stale rows behind.
What happens if a nightly load fails?
Chunked batch processing records checkpoints, and a failed run replays from the exact step that broke. You fix the cause and replay, rather than rebuilding the table by hand.
Can we rerun history after changing the mapping?
Yes. Bulk replay reruns a set of past runs, and what-if replay lets you preview how a change would have handled a recorded run before you promote it.
Related use cases
Lead Routing and Enrichment
Enrich new Typeform submissions with Clearbit firmographics, score them, and route each lead to the right Salesforce owner within minutes.
Customer SuccessUsage-Based Health Scoring
Roll product usage from Snowflake into a Gainsight health score on a schedule and alert the account team in Slack when an account trends down.
AgenticCustomer Feedback Synthesis Agent
An agent gathers feedback from Zendesk tickets, app store reviews, and G2 reviews, finds recurring themes, and sends a cited digest to leadership.
Want this running on your stack?
Neblex Integration Fabric is in beta: full platform, free while in beta. Bring this workflow and we will map it to your systems.