What’s Changed: 2020 Gartner Magic Quadrant for Data Science and Machine Learning Platforms

What’s Changed: 2020 Gartner Magic Quadrant for Data Science and Machine Learning Platforms

The editors at Solutions Review highlight what’s changed since the last iteration of Gartner’s Magic Quadrant for Data Science and Machine Learning Platforms and provide analysis of the new report.

Analyst house Gartner, Inc. recently released the 2020 version of its Magic Quadrant for Data Science and Machine Learning PlatformsGartner defines this software space as enabling “expert data scientists, citizen data scientists and application developers to create, deploy and manage their own advanced analytic models.” The market is seeing rapid changeover as vendors incorporate a wide array of technologies and product capabilities into their platforms. Data science and machine learning software incorporates things like data ingestion, data preparation, data discovery and advanced modeling, testing and deployment.

The researcher notes that machine learning is the most popular singular focus for these platforms, as many buyers need assistance with building data science models. Gartner recommends that each provider be assessed for machine learning independently. As a result, products that are not cohesive and mostly use bundle packages and libraries are not considered true data science and machine learning platforms and are omitted from this analysis. So too are open-source platforms, as the researcher chose only to include those vendors with commercially licensable products.

The data science market features perhaps the most diverse range of solution providers of any data-centric industry. This is definitely purposeful, as use cases and requirements vary from organization to organization with some needing a more augmented approach via machine learning to build and model. Other providers offer more advanced functionality aimed at expert users. However, Gartner notes that the augmented approach may begin to gain steam there as well so data scientists can more efficiently navigate the model-building process.

In Gartner’s report notes, they highlight that market leaders “may not be the best choice.” We applaud this sentiment, as it means that vendor placement isn’t the most important thing readers should take away. The diversity of solution providers and the capabilities offered make it more important than ever for buyers to consider only the tools and products that best line up with their environments and end-goals. It’s true too that different users (citizen data scientist, line-of-business, corporate teams) are likely to have different needs, and this is a key differentiator for a number of the products covered in the study.

Gartner adjusts its evaluation and inclusion criteria for Magic Quadrants as software markets evolve. As a result, Altair (Datawatch) has been added to the report. SAP was removed because it recently launched a new data management and machine learning product for which Gartner needs more time to analyze.

In this Magic Quadrant, Gartner evaluates the strengths and weaknesses of 16 providers that it considers most significant in the marketplace, and provides readers with a graph (the Magic Quadrant) plotting the vendors based on their ability to execute and their completeness of vision. The graph is divided into four quadrants: niche players, challengers, visionaries, and leaders. At Solutions Review, we read the report, available here, and pulled out the key takeaways.

Alteryx is the overall leader after being positioned as a challenger in 2019. Its vertical standing for ability to execute is the highest of any provider in the space. Gartner analysts rank Alteryx’s product vision for augmented data science and process automation as the best in the marketplace. Alteryx acquired automated data preparation company ClearStory Data in April. SAS features a number of enterprise-ready data science tools and recently added new AI and machine learning capabilities to its analytics platform. Like Alteryx, SAS saw an upgraded standing on Gartner’s vertical axis. SAS also offers “one of the best” model operationalization and management platforms right now.

Databricks made the visionaries-to-leaders leap on the heels of its Unified Analytics Platform, which enables customers to keep pace with open-source framework and library releases via ML Runtime. Databricks is making chunk gains in enterprise adoption with reference users praising the provider’s customer success engineers as a major value-add. Databricks raised $650 million in new funding in 2019 (1, 2). TIBCO retains its position as a leader partly due to excellent support, as reference customers speak highly of implementation and ongoing services. The provider is constantly integrating functionality from its various analytics products as well. Users also have access to a complete, end-to-end platform and top-notch edge analytics and streaming IoT capabilities via TIBCO Data Science.

Dataiku and MathWorks are 2020 leaders after being upgraded from the challengers and visionaries brackets, respectively. Dataiku touts ease of use and the ability to support multiple user types in collaboration. Its platform can serve an array of use cases as well. Dataiku 6 was made available in December with new visual features and white box AI. MathWorks is an established name in the engineering and scientific community, and owns one of the top user communities of any vendor in this category. MathWorks offers a fully integrated product, prebuilt solutions (via user-solved problems), and a completely engineered and open environment for the industries it plays in.

IBM is the lone representative in Gartner’s challengers bracket this year, though the mega-vendor did make a fairly dramatic move from market visionary in 2019. IBM is best for multicloud data science via its Cloud Pak for Data. Reference customers speak highly of IBM’s service and support offerings, and Watson Studio now offers new and enhanced collaboration features like catalogs, asset galleries, asset lineage, and permission management. However, product bundling and configuration remains a key issue for a provider with so many data science tools at its disposal.

Microsoft holds the top spot in the visionaries column. It is one of the only vendors included in this report to offer tools for every data science user persona. Microsoft is also one of the fastest providers when it comes to product releases, upgrades, and enhancements. Azure ML performance and scalability scores are among the highest in the industry as well. DataRobot’s augmented data science and machine learning product features tools for both citizen and expert data science professionals. It touts a growing network of partnerships, integrations, and a burgeoning user community as well. DataRobot also raised $206 million in Series E funding and acquired Paxata, ParallelM, and Cursor in 2019.

Though KNIME didn’t retain its leader status from last year’s report, its new position ahead of Google in the upper-left portion of the visionaries graphic has its arrow pointing up. KNIME offers an open-source platform (KNIME Analytics Platform) and commercial extension called KNIME Server that includes advanced functions for collaboration, automation, and operationalization. KNIME’s focus in product innovation and its connections in the data science industry make it a worthwhile consideration. A 2019 leader, RapidMiner offers an expansive portfolio of data science and machine learning tools. RapidMiner Studio supports a range of technologies and its new focus on enterprise organizations means we probably haven’t seen the end of its status as a major player.

H2O.ai and Domino Data Lab round out the visionaries column. H2O.ai offers high performance machine learning and an eye toward emerging trends like augmented data science and explainability. H2O earned the top score in completeness of vision on Gartner’s horizontal axis for 2020. The provider has also begun to branch out to include vertical-specific solutions through partners. H2O raised $72 million in funding last year. Domino Data Lab greatly improved its completeness of vision this year and offers an “industrial-strength, feature-rich and tools-agnostic platform” for data science on-prem or in the cloud. Gartner reference customers mostly choose Domino to improve business process outcomes and support collaboration between business users and data science teams.

Anaconda remains a niche player as its Anaconda Enterprise is best suited for data scientists who use Python or R. The vendor’s user community is a major driver of its product enhancements and user support, though there’s also an enterprise support desk and professional services branch. Anaconda recently improved scalability to serve more users, as well as better support for security and GPU deployments. Altair is the final representative vendor included in Gartner’s Magic Quadrant. Now known as the Altair Knowledge Studio (formerly Datawatch, acquired in 2018) features ease of use and an appeal for coders and non-coders alike. Reference customers (new and inherited) report excellent account management, integration and deployment from the provider.

Read Gartner’s Magic Quadrant for Data Science and Machine Learning Platforms.

Follow Tim

Timothy King

Senior Editor at Solutions Review
Timothy is Solutions Review's Senior Editor. He is a recognized thought leader and influencer in enterprise BI and data analytics. Timothy has been named a top global business journalist by Richtopia. Scoop? First initial, last name at solutionsreview dot com.
Timothy King
Follow Tim