There’s an interesting truth about analytics on big data—once you complete your first analysis and find those deep insights you always thought required a CS or statistics degree from a top-tier school, you just want more.
As businesses become more data-driven, employees in more departments need data to manage their universe. Not static bar charts and pie charts, but real, raw, granular data that they can explore and visualize themselves. People are realizing that the data pipelines and BI workflows that supported them for decades no longer cut it.
This is creating a trend toward a “data-native” approach to analytics that provides BI and analytics professionals direct access to data within modern data platforms such as Hadoop and the cloud.
The benefits of visually analyzing data directly within the data platform are significant, as captured in the recent report from research firm Ovum, “On the Radar: Arcadia Data Unifies Visual Analytics and BI on Hadoop.
A new wave of vendors are looking to extend self-service analytics to big data, going well beyond the pioneering capabilities of vendors like Tableau. Here are a few of the highlights related to a data-native approach to analytics from the report:
- Ultimately, enterprises need to deliver “high-concurrency, real-time access to the full corpus of big data.” The data-native approach addresses the challenges that appear when enterprises want to combine concurrency with scale.
- A data-native approach converges data discovery, BI, and data visualization in a single tool, “functions that normally require multiple tools or the creation of predefined OLAP cubes within Hadoop.” A data-native solution provides analysts access to granular, detailed data directly within the data platform, and IT finds that the layers and complexities of the BI stack are greatly reduced.
- The data-native approach simplifies management, reduces data movement, and yields other benefits not available with traditional approaches.
All of these are important considerations for enterprises exploring—and investing in—the future of analytics, and they provide the basis for what we have developed in Arcadia Enterprise. In the report, Ovum directly calls out that Arcadia “delivers analytics at scale and can do so for large populations of concurrent users.” As they state, the data-native approach allows you to converge the functions of a variety of tools, which eliminates much of the complexity of the traditional data and BI stack. In our case, Arcadia Enterprise “works directly with Hadoop features, such as the security already enforced by the platform; system management, such as Cloudera Manager or Apache Ambari; the interactive SQL query engines (e.g., Impala, Spark SQL, Drill) packaged with the vendor’s platform; and non-relational engines such as Solr and HBase.”
As an aside not directly related to the data-native approach, we appreciate that Ovum recognized our technology, “which greatly increases the addressable audience for big data visualizations.”
Download the complete report for free.
If you’re interested in a more technical overview of the evolution from cube-based BI to the data-native approach to analytics, check out our blog post, Beyond the Cube: Embrace Analytical Views. The gist of that article is that if you go the traditional route of creating a fast analytical environment for your users, you spend a lot of time creating OLAP cubes. That not only puts a strain on your IT team, but it also delays your users’ access to data. Analytical Views are an optimized data structure to provide quick query results in analytical applications without the time-consuming planning and analysis that OLAP cubes require.