Snowflake Marketplace and Publishing Data Products
The Snowflake Marketplace is a global exchange where organisations publish live datasets and data services for others to discover and use directly inside their own Snowflake accounts. Thousands of providers list datasets covering weather, financial markets, demographics, geospatial data, healthcare benchmarks, consumer sentiment, and much more. You subscribe to a dataset and query it with SQL immediately — no file downloads, no data pipelines, no API integrations.
Why the Marketplace Changes How Organisations Use Data
Before the Snowflake Marketplace, enriching your internal data with external data required a long procurement and engineering process: find a vendor, negotiate a data license, receive a file or API credential, build a pipeline to ingest the data, schedule regular refreshes, and monitor for changes in file format or schema. This process typically took weeks and required ongoing engineering maintenance.
With the Marketplace, that entire process collapses to three clicks. You find a dataset, click Get, and it appears in your account as a live database. The provider updates it on their side; your queries always return the latest version. You contribute no storage cost because the data lives in the provider's account.
OLD EXTERNAL DATA WORKFLOW SNOWFLAKE MARKETPLACE WORKFLOW ------------------------------- ------------------------------- 1. Find vendor (days) 1. Search Marketplace (seconds) 2. Sign license agreement (weeks) 2. Click "Get Data" (1 click) 3. Receive data file (days) 3. Select warehouse (1 click) 4. Build ingest pipeline (weeks) 4. Query live data (immediate) 5. Schedule daily refresh (days) 6. Monitor for schema changes (ongoing) 7. Rebuild pipeline when format changes Total time to first query: weeks-months Total time to first query: < 5 minutes
Navigating the Marketplace
Open the Marketplace section from Snowsight's left sidebar. The home page shows featured listings and category tiles. The search bar at the top lets you search by keyword, provider name, or category. Each listing shows the provider name, dataset name, a description, pricing (free or paid), and the regions where the data is available.
MARKETPLACE CATEGORY OVERVIEW
==============================
CATEGORY EXAMPLE PROVIDERS AND DATASETS
-------- ------------------------------
Financial Data Bloomberg market data, Refinitiv, Intrinio stock prices
Weather The Weather Company, NOAA historical weather,
AccuWeather forecasts
Geospatial SafeGraph foot traffic, HERE maps, US Census boundaries
Demographics Acxiom consumer profiles, Nielsen audience data
Healthcare IQVIA prescription data, CMS Medicare claims benchmarks
Business Intel. Dun & Bradstreet company data, Kompass B2B contacts
Marketing LiveRamp identity graph, DoubleVerify brand safety
Supply Chain Panjiva import/export records, Datamyne trade data
Sustainability Urgentem carbon emissions, CDP climate scores
Accessing a Free Marketplace Dataset
Most marketplace listings have a free tier. The steps below walk through accessing a real free dataset — weather data from NOAA (National Oceanic and Atmospheric Administration) provided through a Snowflake partner.
STEP-BY-STEP: GET A FREE MARKETPLACE DATASET
=============================================
Step 1: Open Marketplace in Snowsight left sidebar
Step 2: Search for "Weather" in the search bar
Step 3: Click on a listing (e.g., "Weather Source LLC: Global Weather")
Review the listing:
- Provider: Weather Source LLC
- Coverage: Global daily weather observations
- Update frequency: Daily
- Sample queries: yes (shown in listing)
- Price: Free (basic tier)
Step 4: Click "Get" button (top right of listing)
A dialog appears asking:
- Database name in your account: type "WEATHER_DATA" (or any name)
- Which roles can access it: select SYSADMIN and PUBLIC
Click "Get Data"
Step 5: Snowflake creates a shared database "WEATHER_DATA" in your account
This takes 10-30 seconds
Step 6: Query the data immediately
USE DATABASE WEATHER_DATA;
SHOW SCHEMAS; -- see available schemas
SHOW TABLES; -- see available tables
SELECT *
FROM WEATHER_DATA.STANDARD_TILE.HISTORY_DAY
WHERE POSTAL_CODE = '10001' -- Manhattan, New York
AND DATE_VALID_STD >= '2024-01-01'
ORDER BY DATE_VALID_STD
LIMIT 30;
Combining Marketplace Data with Your Own Data
The real power of Marketplace data emerges when you JOIN it with your own business data. Your sales data plus external weather data can reveal whether bad weather affects in-store sales. Your supply chain data plus external port congestion data can predict shipping delays.
-- Enrich internal sales data with external weather data
SELECT
s.store_id,
s.store_city,
s.sale_date,
s.daily_revenue,
w.avg_temperature_air_2m_f AS avg_temp_f,
w.tot_precip_in AS precipitation_inches,
w.avg_wind_speed_100m_mph AS wind_speed_mph,
CASE
WHEN w.tot_precip_in > 0.5 THEN 'Rainy'
WHEN w.avg_temperature_air_2m_f < 32 THEN 'Freezing'
WHEN w.avg_temperature_air_2m_f > 90 THEN 'Hot'
ELSE 'Mild'
END AS weather_condition,
CORR(s.daily_revenue, w.avg_temperature_air_2m_f)
OVER (PARTITION BY s.store_id) AS revenue_temp_correlation
FROM RETAIL_DB.ANALYTICS.STORE_DAILY_SALES s
LEFT JOIN WEATHER_DATA.STANDARD_TILE.HISTORY_DAY w
ON s.store_postal_code = w.postal_code
AND s.sale_date = w.date_valid_std
WHERE s.sale_date >= '2024-01-01'
ORDER BY s.store_id, s.sale_date;
Paid Marketplace Listings
Premium datasets come with a subscription fee. The listing shows the pricing model clearly — flat monthly fee, per-query pricing, or tiered access based on how much data you consume. You subscribe through the Marketplace interface using your Snowflake billing account. The provider receives payment through Snowflake; you do not need a separate contract with the data provider for standard listings.
PAID LISTING PRICING MODELS
=============================
MODEL HOW IT WORKS BEST FOR
----- ------------ --------
Flat monthly fee Fixed cost regardless of usage Heavy users of that dataset
Per-query pricing Pay only for queries you run Occasional/exploratory use
Tiered access More data fields at higher tiers Starting small, growing later
One-time purchase Pay once for a static dataset Historical snapshot data
Bring your own key You negotiate with provider Enterprise-scale agreements
outside Marketplace
Publishing Your Own Data to the Marketplace
If your organisation owns valuable data, you can publish it as a Marketplace listing and generate revenue from it — or share it privately with selected partners. The process requires a Snowflake Business Network account and approval from Snowflake.
PUBLISHING WORKFLOW
====================
Step 1: Prepare your data
- Create clean, well-documented tables or views
- Write sample queries that demonstrate value
- Define which columns/rows to expose (usually via secure views)
Step 2: Create a Listing in Provider Studio
- Open Marketplace > Provider Studio
- Click "Create Listing"
- Set visibility: Public, Private, or Personalized
Step 3: Configure the listing
LISTING SETTINGS:
Title: Descriptive, searchable name
Description: What the data contains, how it is collected, update frequency
Categories: Pick 1-3 relevant categories for discoverability
Sample queries: Include 3-5 ready-to-run example queries
Data dictionary: Document every column name and meaning
Regions: Which cloud regions support the listing
Pricing: Free / Flat fee / Usage-based
Step 4: Add the data share to the listing
- Choose an existing Share or create a new one
- The Share controls which tables/views consumers receive
Step 5: Submit for review
- Snowflake reviews the listing for content quality and compliance
- Approval typically takes 1-3 business days
Step 6: Listing goes live
- Public listings become searchable by all Snowflake users globally
- Private listings are visible only to invited accounts
Personalised Listings and Data Clean Rooms via Marketplace
Personalised listings let a provider publish a listing where each consumer receives a custom view of the data. The provider configures row-level filters per consumer account — for example, a weather provider might give each airline only the weather data for airports that airline flies into, rather than global data.
PERSONALISED LISTING DIAGRAM
==============================
WEATHER PROVIDER
|
|-- Listing: "Airport Weather Premium"
|
+--> Airline A's account:
| Receives only: airports in North America
|
+--> Airline B's account:
Receives only: airports in Europe and Asia
Configuration uses row access policies or parameterised secure views:
CREATE SECURE VIEW VW_AIRPORT_WEATHER AS
SELECT *
FROM AIRPORT_WEATHER_GLOBAL
WHERE airport_region = CURRENT_ACCOUNT_LOCATOR();
-- Snowflake substitutes each consumer's account locator automatically
Monitoring Marketplace Data Usage
-- As a provider: see which accounts query your shared data
SELECT
consumer_account_locator,
query_date,
COUNT(*) AS query_count,
SUM(bytes_scanned) AS bytes_scanned
FROM SNOWFLAKE.ACCOUNT_USAGE.DATA_SHARING_USAGE
WHERE share_name = 'MY_MARKETPLACE_SHARE'
AND query_date >= DATEADD('day', -30, CURRENT_DATE())
GROUP BY 1, 2
ORDER BY 2 DESC, 3 DESC;
-- As a consumer: see which marketplace databases you have
SHOW DATABASES LIKE 'WEATHER%';
SELECT * FROM SNOWFLAKE.ACCOUNT_USAGE.DATABASE_STORAGE_USAGE_HISTORY
WHERE database_name = 'WEATHER_DATA';
-- Storage shows 0 bytes — confirms no data copy in your account
Key Points
- The Snowflake Marketplace lists thousands of live datasets from external providers — subscribe and query in minutes without any data pipeline work
- Subscribed datasets appear as shared databases in your account and consume zero storage — data stays in the provider's account
- Combine Marketplace data with your own tables using standard SQL JOINs — weather, demographics, or financial data enriches your internal datasets instantly
- Paid listings support flat monthly fees, per-query pricing, and tiered access — billing happens through your existing Snowflake account
- Publish your own data to the Marketplace through Provider Studio — control visibility (public or private) and define custom secure views per consumer
- Personalised listings allow row-level customisation per consumer account, letting one listing serve different audiences different slices of the same dataset