Databricks Marketplace is an open marketplace for data, analytics, and AI, powered by the open-source Delta Sharing standard. Since the release of Databricks Marketplace, we have seen 300% growth in listings and providers. Throughout this explosive growth phase, we helped data partners and data consumers make the best use of Databricks Marketplace.
Today, we want to share the top 10 most frequently asked questions. Tailored for both consumers and providers, read on to learn more about what makes us different, how to make your listing stand out, and how to find the right product for your specific data needs.
Why should I use Databricks Marketplace as a consumer?
Databricks Marketplace includes a variety of products such as data sets, ML models, notebooks, solution accelerators, and very soon, applications. This variety supports a broader range of analytics and AI initiatives, helping consumers realize the full value of the data. There are 2500+ listings across 250+ providers.
Databricks Marketplace allows for data access without the need to be on the Databricks platform. This open approach avoids vendor lock-in and enables consumers to use their preferred tools and platforms, enhancing flexibility and integration.
Built on top of Delta Sharing, Marketplace eliminates the need for complex ETL processes and expensive data replication. Customers can use Unity Catalog to govern the Databricks Marketplace datasets along with the rest of the lakehouse data. All this reduces the operational burden on consumers, allowing them to focus on deriving value from the data rather than managing and governing it.
What makes Databricks Marketplace different from other data marketplaces?
Databricks Marketplace is powered by Delta Sharing, so you can benefit from open source flexibility and no vendor lock-in. It offers a diverse range of assets beyond just structured data, including AI models, notebooks, volumes or unstructured data (images/videos/audio), and solution accelerators, accessible through various analytics tools such as PowerBI and Spark. Examples include Crisp (Tables), Shutterstock (Volumes), AI21 Lab (AI Models), Datavant or LiveRamp (Solution Accelerators), and Kythera Labs (Notebooks).
D2O (Databricks-to-Open) sharing allows Databricks Marketplace providers to share data with recipients on any platform or cloud, without requiring you to be on Databricks. This expands the addressable market for providers, while deep integration with Unity Catalog enhances governance for Databricks users. These features collectively emphasize flexibility, openness, and comprehensive support for diverse data and analytics needs, catering to a wide range of users across different platforms and tools.
How do I figure out what data product is right for my needs? How can I evaluate this comprehensively?
To identify the most suitable data product on Databricks Marketplace, begin by specifying your analytical goals and data requirements. Utilize the platform’s filtering capabilities to refine your search by product, provider, category, model task, and type of listing. For an in-depth evaluation, take advantage of the pre-built notebooks and sample datasets available; these resources enable you to conduct exploratory data analysis and assess the data’s fit for your specific use case.
For example, the HealthVerity Claims Sample Patient Dataset includes a detailed notebook and use cases to assist you in assessing its applicability. Technical users can execute sample queries within these notebooks to gain insights into the data’s schema, quality, and potential analytical applications.
Once I’m interested in a data product listed on the Marketplace, how do I get access to it?
Data and AI products are offered through two types of listings – instantly available and request for access. For listings that are instantly available, you can simply click the “Get instant access” button on the listing page, agree to the terms and conditions, and immediately access the data. This type of listing is typically used for free or public datasets and does not require any approval from the provider. See all the free listings here.
For request for access listings, you can click the “Request access” button, provide your basic information and intended use, and wait for provider approval. Once the provider has completed reviewing your request, you will be notified via email and receive instructions on how to access the data. This type of listing is typically used when the data involves a commercial transaction, customization for specific needs, or other business agreements that need to be finalized outside of the Marketplace platform.
Examples of free listings include AccuWeather Historical Weather Data Set by Zip Codes, and request-for-access listings include IQVIA PharMetrics® Plus Claims Data.
What are private exchanges and when should I use them?
Leverage the private exchange feature to use Databricks Marketplace for private data sharing. This allows you to make data and AI products discoverable only to a specific group of consumers. With Private Exchanges, you can facilitate data sharing across different subsidiaries or departments within a large organization, or across different organizations, using a marketplace-like interface for discovery and fulfillment. For more details, refer to this blog post.
Why should I use Databricks Marketplace as a provider?
We know there are multiple marketplace options when it comes to listing your data products. However, Databricks Marketplace stands out by maximizing the reach of provider’s data across platforms, regions, and clouds without enforcing proprietary systems. So, we are not forcing providers or recipients to adopt a proprietary platform. Secondly, Databricks reduces TCO. The platform’s Delta Sharing protocol enables data providers to share live data directly from their Databricks environment without needing replication. Lastly, providers can share AI Models with Databricks Marketplace, opening up new revenue streams.
Check out how Deutsche Börse streamlined data sharing across cloud environments, resulting in 2-4x faster insights and 4-8x faster publishing. Allium, another Databricks Marketplace provider, leveraged Delta Sharing with Cloudflare R2, saving $645K annually and transferring 1PB of data monthly.
How do I become a provider and list my data and AI products assets on Databricks Marketplace?
There are two paths to becoming a provider. To create public listings, simply apply through the Databricks Data Partner Program. Once approved, you will gain access to the provider console in your Unity-Catalog-enabled Databricks workspace. If you prefer to share data only with select Databricks recipients (either cross-account or internal) and are a Marketplace admin, you can sign up directly through the provider console as a private exchange provider.
How do I reach new customers and commercialize my listings on Databricks Marketplace?
We understand that data product sales are often highly customizable and shouldn’t be forced into a standardized out-of-the-box pricing model. To allow for that flexibility, all commercial transactions take place directly between you and the consumer.
To ensure your data products on Databricks Marketplace are effectively reaching potential customers, leverage the newly released Provider Analytics Dashboard. This tool is essential for monitoring consumer activity and understanding which of your listings are generating the most interest. By analyzing metrics from the dashboard, you can identify promising leads and initiate personalized conversations about licensing and commercialization. This approach allows for a flexible, tailored sales process that aligns with the customizable nature of data product transactions, facilitating effective lead generation and customer engagement.
How do I make my listing stand out?
The more clarity and color you can paint with your listing, the more it’ll attract the right consumers! This means having detailed and accurate information in each listing field and using category tags and attributes to enhance discoverability. We’ve also found that listings with sample notebooks have more success by allowing consumers to interact directly and see sample queries in action. Offering trial versions of the data can also entice potential customers to explore and experience the value of your data firsthand, boosting engagement and conversion rates.
Beyond that, it’s always a good idea to supplement your listing with educational resources such as demos, webinars, tutorials, or blog posts that guide users through the process of using your dataset effectively. Create demonstrations that showcase how data engineers can easily acquire and integrate your data into their existing systems or build demos that show how data scientists can use your datasets to train machine learning models.
Some notable listings include Corelogic for its comprehensive notebook, Zoominfo for its detailed explanation of use cases and inclusion of a customer case study, and AI21 Labs for its extensive notebook showcasing advanced AI models, along with a blog.
Is Databricks Marketplace free? Is there a fee to list on Databricks Marketplace?
There is no fee for providers to list their assets on Databricks Marketplace. We host both free and paid listings, but we don’t take a cut of the revenue that providers receive. Transactions for paid listings happen between consumers and providers directly.
The Databricks Marketplace is revolutionizing how organizations access and leverage data, AI models, and analytics solutions. Don’t miss out on the opportunity to transform your business – explore the Databricks Open Marketplace today and unlock the power of sharing and collaboration! If you don’t find what you’re looking for on Databricks Marketplace, let us know!
Source link
lol
