Google Cloud Storage in Polar: drop a file, get a governed metric.

Land the datasets your team already exports to a GCS bucket straight into Polar. Cost files, offline and wholesale orders, logistics exports and proprietary KPIs become governed metrics and dimensions, blended with Shopify, Meta and Google in one semantic layer your team and your AI can query. No pipeline to build, no reverse-ETL, no SQL to write in Polar.

/ ⏎

Connect Google Cloud Storage See the setup

Google Cloud Storage → Polar

The Google Cloud Storage files your team already exports, live next to Shopify and ads.

Point Polar at a bucket, drop your CSV or Parquet files in, and their columns become queryable metrics and dimensions blended with revenue, spend and orders. No manual uploads, no duplicate pipeline, refreshed on every scheduled sync.

Custom Datasets

“Can I just use the files we already export?”

Any file you drop in your GCS bucket can be mapped into Polar's semantic layer, so its columns turn into governed metrics and dimensions. The export you already run carries over, not the maintenance.

bucket files -> metrics

Finance & Cost

“Where do COGS and landed cost live?”

Drop your cost files (COGS, landed cost, overheads, fees) into the bucket and Polar blends them with Shopify revenue and ad spend for real contribution margin, not gross revenue.

COGS & landed cost

Custom KPIs

“My KPI isn't in any standard connector.”

Whatever your team computes and exports (custom LTV, cash conversion, unit economics) becomes one governed Polar metric everyone reads the same way.

custom KPI, governed once

Offline & Wholesale

“What about revenue that never touches Shopify?”

Wholesale, retail, marketplace and offline order exports land in the bucket and sit next to DTC, so the P&L reflects the whole business, not just the storefront.

offline + DTC, one P&L

Logistics & Ops

“Can I bring in shipping and 3PL data?”

Logistics exports, 3PL fees, inventory snapshots and delivery data become dimensions and metrics you can slice revenue and margin by.

logistics & 3PL exports

Blended Context

“How does file data sit with everything else?”

Your dropped files live in the same semantic layer as Shopify, Meta, Google and Klaviyo, blended into Net Sales, MER, CAC and margin, and readable by Ask Polar and the MCP.

files in the blended layer

Setup Guide

How to Connect Google Cloud Storage to Polar.

A Polar account with Shopify connected, read access to the GCS bucket where you drop your files, and the datasets you want to expose. Polar's team maps them into your semantic layer. Read-only is enough, Polar never writes to your bucket.

01 · What you need

A Polar account with Shopify

Google Cloud Storage plugs in next to your store data, so your exported files meet real revenue and spend automatically.

Connect your data →

Read access to your GCS bucket

A service account with read access to the bucket and folders you want in Polar. Read-only is enough, Polar never writes back to your storage.

Google Cloud Storage →

The files you want to expose

Pick the exports, cost files and custom-KPI datasets to bring in.Google Cloud Storage sits under Custom connections available in Connectors. Click Request and Polar's AI Data Engineer builds it, in minutes.

02 · Connect it, the guided way

Request Google Cloud Storage in Connectors ~1 min

Custom connections available.

In Polar, open Connectors. Google Cloud Storage sits under Custom connections available: click Request and Polar's AI Data Engineer builds it, in minutes.

Open Polar connectors

Grant read access and pick your files ~5 min

A read-only service account and the datasets to expose.

Create a read-only service account with access to the bucket and folders you drop files into, and share the connection details.

Name the files and custom datasets to bring in. Polar maps each into your semantic layer as governed metrics and dimensions, refreshed on your next scheduled sync.

Best practice: start with the one or two files you report on most (a COGS export, a custom-KPI file). Drop new versions in the same path and Polar picks them up automatically, no manual upload.

See it in your reports the payoff

File data, blended and governed.

Your exported files now power dashboards, reports and every Ask Polar answer, blended with Shopify, Meta and Google. They feed contribution margin, true CAC and MER, with no SQL to write in Polar.

Side by side

Google Cloud Storage stores the data. Polar turns it into governed metrics.

Files land in Google Cloud Storage and stay files: no metric definitions, no blend with your store, and every question needs someone to write the query. Polar ingests them into the same semantic layer as your orders and your spend.

Google Cloud Storage on its ownon its own

Google Cloud Storage with Polarblended

Data sits as files and tables, with no metric layer on top.

Ingested into Polar's semantic layer as governed metrics.

Blending it with orders and spend means writing SQL.

Blended with Shopify, ads and email without writing SQL.

Every new question is a new query and a new export.

Ask in plain language, or read it in a dashboard that refreshes.

No shared definition, so two teams arrive with two numbers.

One definition, so the number is the same everywhere.

Featured Use Case

The true-margin P&L your file exports already built.

Your cost and offline-order files live in a GCS bucket. Blend them with Shopify revenue and ad spend in Polar and every channel gets a real contribution margin, not a gross-revenue guess. No manual upload, no reverse-ETL, no SQL in Polar.

Contribution margin by channel● COGS from GCS

Channel

Revenue

COGS + spend

Margin

Meta paid

$182k

$146k

20%

Email / SMS

$94k

$51k

46%

Offline / wholesale

$120k

$101k

16%

COGS and offline orders dropped as files in GCS, blended with Shopify and ads in Polar.

Mapped once

No AI needed

Map your GCS cost file into the semantic layer once. Contribution margin then shows up in every dashboard, report and scheduled export automatically. Drop a fresh file and it refreshes, no SQL, no rebuild.

AI agent

Dynamic

Give an agent the Polar MCP. Ask it to rank channels by true contribution margin using your bucket COGS, and it queries the blended model in plain language, no SQL required.