I'm building a data pipeline from Salesforce and Stripe into one warehouse

Great use case. A Salesforce + Stripe warehouse is a very common setup for combining revenue, customer, and billing data.

A practical way to approach it:

1) Define the core entities

You’ll usually want to model these as the main shared objects:

• Customer / Account
  • Salesforce: Account, Contact
  • Stripe: Customer
• Opportunity / Deal
  • Salesforce: Opportunity
• Subscription / Contract
  • Stripe: Subscription, sometimes Invoice
• Payment / Invoice
  • Stripe: Invoice, Charge, PaymentIntent
• Product / Price
  • Stripe: Product, Price
• Activity / Lifecycle events
  • Salesforce tasks/calls/emails
  • Stripe subscription/invoice events

2) Decide on the warehouse modeling approach

A good default is:

• Raw/Staging layer
  • Land each source table/event mostly as-is
  • Keep source-specific IDs and timestamps
• Normalized/core layer
  • Standardize naming, types, and keys
  • Build shared dimensions like dim_customer, dim_product, dim_date
• Analytics/mart layer
  • Revenue, ARR, churn, pipeline, conversion, LTV, etc.

3) Key integration challenge: identity matching

You’ll need a clear way to connect Salesforce records to Stripe records.

Typical strategies:

• Email match between Salesforce Contact and Stripe Customer
• Explicit mapping field
  • e.g. store stripe_customer_id in Salesforce or salesforce_account_id in Stripe metadata
• Account-level mapping
  • one Salesforce Account ↔ one or more Stripe Customers
• Fallback fuzzy matching
  • use carefully, ideally only for exception handling

Best practice: create an identity resolution table like:

• entity_type
• source_system
• source_id
• canonical_customer_id
• match_method
• confidence

4) Handle time and state correctly

Stripe especially is event-driven:

• subscription changes
• invoice finalization
• charge succeeds/fails
• refunds
• disputes

Salesforce is more state-oriented:

• current account/opportunity/contact state
• changes over time via history objects or CDC if available

Use:

• incremental syncs
• updated_at / created_at watermarks
• event timestamps vs ingestion timestamps
• slowly changing dimensions if you need history

5) Common warehouse outputs

Useful business tables/metrics:

• Pipeline
  • open pipeline by stage
  • stage conversion rates
  • sales cycle length
• Revenue
  • MRR / ARR
  • booked vs collected revenue
  • recurring vs one-time
• Billing health
  • failed payments
  • overdue invoices
  • churn / downgrade
• Customer health
  • account age
  • support activity
  • expansion/contraction
• Sales ↔ billing linkage
  • opportunity won date vs first invoice date
  • rep attribution for billing events

6) Tooling pattern

Common stack:

• Ingestion: Fivetran / Airbyte / Stitch / custom API jobs
• Transform: dbt
• Warehouse: Snowflake / BigQuery / Redshift / Databricks SQL
• Orchestration: Airflow / Dagster / Prefect

7) A good minimal schema

If you want to keep it simple initially:

Staging

• stg_salesforce_accounts
• stg_salesforce_contacts
• stg_salesforce_opportunities
• stg_stripe_customers
• stg_stripe_subscriptions
• stg_stripe_invoices
• stg_stripe_charges

Core

• dim_customer
• dim_account
• dim_product
• fact_opportunity
• fact_subscription
• fact_invoice
• fact_payment

Mart

• fct_arr
• fct_mrr
• fct_pipeline
• fct_customer_lifetime_value

8) Questions to answer before building

1. What warehouse are you using?
2. Do you need real-time or daily batch is fine?
3. Is Salesforce your source of truth for accounts/customers, or Stripe?
4. Do you need historical state tracking or just current snapshots?
5. How do you want to match Salesforce people/accounts to Stripe customers?

If you want, I can help you design:

• a data model
• an ELT pipeline architecture
• a dbt schema
• or a field mapping between Salesforce and Stripe