Skip to main content

Listings

A well-designed listing helps potential customers understand your data product and get started quickly. This page covers provider profile setup, listing types, content best practices, and analytics. For step-by-step instructions, see List your data product in Databricks Marketplace.

Provider profile​

Your provider profile represents your public identity on the Marketplace. Ensure it is complete, accurate, and professional.

When creating your profile, pay close attention to two fields:

FieldConsideration
Provider nameCannot be edited after saving. To change it later, you must create a new profile and migrate listings.
Business emailWhere all consumer requests are sent. Use a team distribution list, not an individual address.

Listing types​

Choose between public Marketplace listings and Private Exchanges based on your distribution goals.

Public Marketplace​

Public listings appear on the open Databricks Marketplace and are searchable by all Databricks users.

Best for:

  • Broad market exposure and brand recognition
  • Attracting new customers who don't have an existing relationship
  • Data products suitable for general availability

Private Exchange​

Private Exchange listings are only visible to customers you explicitly allow based on their metastore sharing identifier. They do not appear in public search results.

Best for:

  • Targeted collaboration with specific customers
  • Pre-commercial or early-access data products
  • Controlled distribution with known partners

Listing content​

Each listing should clearly represent the data or assets being shared. Include a descriptive summary that outlines:

  • The dataset's purpose and business value
  • Intended use cases and target audience
  • Key details: coverage, frequency, update schedule, structure
  • Data quality and governance practices

This is your opportunity to tell the story of your data and help potential customers understand how it solves their business problems.

What makes a listing stand out
  • Lead with the business problem you solve, not the data you have
  • Include a concrete use case example ("Predict customer churn with 90-day lead time")
  • Quantify coverage and freshness ("50M+ US businesses, updated weekly")
  • Highlight unique differentiators ("Only dataset with verified mobile numbers")

Metadata and discoverability​

Good metadata improves discoverability for both humans and AI:

  • Use clear, descriptive titles that include the data domain
  • Add relevant categories and tags from the Marketplace taxonomy
  • Include industry and use case keywords in descriptions
  • Document data freshness and update frequency
  • Add rich table and column comments—these power Genie and improve AI-assisted discovery

See AI readiness for comprehensive metadata guidance.

Access models​

Your access model determines the buyer experience—instant for frictionless onboarding, request-based for controlled distribution.

Instant access​

Consumers click Get Instant Access, agree to your terms of use, and receive immediate access with no manual steps required.

Best for:

  • Broadly shareable data
  • Evaluation and trial datasets
  • Low-friction onboarding

Request access​

Consumers submit a request with their name, email, company, and use case. You review and approve requests manually.

Workflow:

  1. Consumer clicks Request Access and submits form
  2. You receive an automated email with request details
  3. Request appears in your Consumer Requests table in the Provider Console
  4. You review and approve, attach the appropriate share, or reject

Best for:

  • Premium or paid data products
  • Data requiring compliance review
  • Custom entitlements per customer

See Access management for fulfillment workflows.

Sample notebooks​

You can add up to ten sample notebooks per listing. These notebooks act as a guided, low-friction onboarding experience—not a showcase of every capability, but a path to the first meaningful query.

Clear purpose and orientation​

Start with a short introduction that answers:

  • What the dataset or product represents
  • Who it is intended for
  • What problems or use cases it supports
  • What the consumer will accomplish by running the notebook

Representative, realistic queries​

Include a small number of high-value example queries that:

  • Reflect common customer use cases
  • Demonstrate realistic joins, filters, or aggregations
  • Execute quickly and deterministically

Best practices:

  • Prefer clarity over cleverness
  • Avoid edge-case logic
  • Avoid long-running or cost-intensive operations

Include visualizations​

Add sample code in SQL, Python, Scala, or R, along with visualizations and common workflows. A well-designed notebook dramatically reduces time to value.

Notebook checklist
  • Clear intro with data overview
  • 3-5 representative queries
  • At least one visualization
  • Runs in under 2 minutes
  • No hardcoded credentials or paths
Avoid these pitfalls
  • Notebooks that require external dependencies or setup
  • Queries that take minutes to run or cost significant compute
  • Stale notebooks that reference outdated schema

Keep notebooks in sync with your data—update examples when schema changes and test before each listing update.

Listing analytics​

Track listing performance using Marketplace system tables.

Key tables​

TablePurpose
system.marketplace.listing_access_eventsCaptures every completed request or "get data" action. Includes listing metadata and consumer details.
system.marketplace.listing_funnel_eventsRecords impressions, clicks, and other user actions. Helps identify drop-offs.

Sample queries​

Access requests by listing (last 30 days):

SELECT 
  listing_name,
  event_type,
  COUNT(*) AS request_count
FROM system.marketplace.listing_access_events
WHERE event_time >= CURRENT_DATE - INTERVAL 30 DAYS
GROUP BY listing_name, event_type
ORDER BY request_count DESC;

Funnel conversion rate:

SELECT 
  listing_name,
  SUM(CASE WHEN event_type = 'VIEW_LISTING' THEN 1 ELSE 0 END) AS views,
  SUM(CASE WHEN event_type = 'START_GET_DATA' THEN 1 ELSE 0 END) AS started,
  SUM(CASE WHEN event_type = 'COMPLETE_GET_DATA' THEN 1 ELSE 0 END) AS completed,
  ROUND(
    100.0 * SUM(CASE WHEN event_type = 'COMPLETE_GET_DATA' THEN 1 ELSE 0 END) / 
    NULLIF(SUM(CASE WHEN event_type = 'VIEW_LISTING' THEN 1 ELSE 0 END), 0), 
    2
  ) AS conversion_pct
FROM system.marketplace.listing_funnel_events
WHERE event_time >= CURRENT_DATE - INTERVAL 30 DAYS
GROUP BY listing_name
ORDER BY views DESC;

Marketplace Dashboard​

The Marketplace Dashboard in the Provider Console provides a high-level view of:

  • Listing views and impressions
  • Access requests and fulfillment rates
  • Active recipients
  • Conversion rates from discovery to access

Use these insights to refine listing content, identify friction points, and improve conversion.

What's next​