How should we connect Shopify data to a data warehouse?
Our answer
Add a warehouse only with several data sources and someone to own it, about past $10M a year. Then use a managed connector, grant every read scope, keep a few days of lookback and reconcile net sales to Shopify.
- Confidence
- Medium
- Strongest proof
- 3 official rules
- 4 practitioners, 5 opinions or claims
- Sources
- 6 sources
- 12 statements
PA
FGA
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentBelow about $10M in sales with few sources, a warehouse is not worth it (Polar Analytics, ecommerce analytics software maker). Eightx (fractional CFO firm) argues buying wins until someone can own the data, usually past $10M to $15M.
Custom scripts need ongoing upkeep as Shopify's API changes, Fivetran argues. You don't need to write them. Fivetran, Airbyte and Google each offer a ready Shopify connector, though Google's is in Preview.
Connector setup errors hide in plain sight. A missing scope shows up as an empty table, and late order edits need a few days of lookback (Airbyte). Shopify allows five bulk queries per app at once.
Our judgment
Wait on a warehouse until a store has several data sources, unanswered questions and an owner for the data. Eightx ties the switch to having that person, and Polar ties it to many sources and three unanswered questions. Use a managed connector rather than custom code, because Fivetran warns of upkeep as Shopify's API changes, Polar names connector maintenance as a hidden cost of building, and Fivetran, Airbyte and Google each offer a ready connector. Check warehouse net sales against Shopify's figures before anyone uses them. That combines Shopify's net sales definition with Polar's point that the orders model for refunds is the hardest part to get right.
Our judgment, built from Eightx, Polar Analytics, Fivetran, Airbyte, Google and Shopify: our reading of these sources, not a finding from any one of them.
What the sources say
12 statements from 6 sourcesFivetran recommends its GraphQL schema for Shopify connections, since Shopify's REST Admin API is in maintenance. Existing connections cannot switch; a new one starts a full initial sync.
“You cannot directly switch an existing connection from the Legacy schema to the GraphQL schema.”
Shopify's docs say from API version 2026-01 each app can run up to five bulk query operations per shop at once, each finishing within 10 days, with results as a JSONL file.
“In API versions 2026-01 and higher, each app can run up to five bulk query operations per shop simultaneously.”
Shopify defines net sales as gross sales minus discounts and sales reversals, excluding taxes, shipping, duties and fees; total sales adds those back. It also tags each order as new or returning.
“Net sales = gross sales - discounts - sales reversals”
Google's BigQuery Data Transfer Service has a Shopify connector in Preview that loads orders, refunds, customers and products on a schedule, free during Preview. It needs a custom Shopify app.
“There is no cost to transfer Shopify data into BigQuery while this feature is in Preview”
Fivetran, which sells a managed connector, compares four ways to move Shopify data into PostgreSQL, saying custom API scripts need ongoing upkeep and self-hosted connectors still need running and fixing.
“Create a custom app in the Shopify admin panel to generate API credentials.”
Airbyte, which sells a Shopify connector, advises a few days of lookback when syncing Shopify orders and refunds to BigQuery to catch late edits. Its one-bulk-operation limit is stricter than Shopify's own page.
“Shopify records are frequently updated after the fact, and a lookback window catches late edits that a strict cursor would miss.”
Polar Analytics, which sells a marketing data warehouse, says skip a warehouse under about $10M GMV with few sources, buy one from $10M to $100M+ without a data engineer, and build only with a real data team.
“A KPI is a definition, not a number, so settle CAC, ROAS, and LTV once.”
Eightx models 24-month reporting costs for a $6M brand from about $12,000 to $32,000 by tool, and says labor drives cost, so buy until someone can own the data.
“The trigger is having the person, not hitting a revenue number.”
Polar Analytics, which sells an analytics platform, says a custom data stack takes 6 to 9 months and $15k to $38k a month before salaries, worth it only with a team and unusual data needs.
“It is the right call only when you have the team and a data model no vendor can express.”
Shopify says its bulk operations now export large datasets up to 4X faster and are the most efficient way to move store data, with up to 5 bulk operations running at once.
“Exporting large datasets from Shopify is now up to 4X faster”
Polar Analytics, which sells a managed Snowflake warehouse, says building one yourself takes 6 to 9 months and a data hire, and brands shouldn't migrate if single-channel or unable to name three unanswered questions.
“Don't migrate if you're single-channel or can't name the three questions your current tool can't answer.”
Eightx, a CFO firm, says DTC brands under $5M in annual sales need no paid attribution tool: platform ROAS reconciled in a spreadsheet is the default. The real cost of any tool is 2 to 6 weeks of tracking cleanup.
“The default DTC stack at every revenue band is platform-native ROAS reconciled in a spreadsheet, not a paid tool.”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer