Becoming a fast, data-driven organization requires a lot of work under the hood. Integrating, processing, and analyzing data from a diverse, endless range of sources—in a way that is cohesive and conducive to business success—remains a prominent challenge for many enterprises.
In DBTA’s webinar, Achieving Unified Analytics: Top Tools and Strategies, experts offered their insights on achieving unified analytics, defined by reimagining traditional data architectures with modern data platforms, tools, and approaches.
Mark Balkenende, global head of product marketing, Incorta, explained that data integration faces a multitude of challenges, including complex, multi-source systems, difficult to build and maintain ETL pipelines, and a need for real-time refreshes. Data management’s traditional issues—including rigid architectures with multiple data hops and multiple copies—further complicate the demand for modern, speedy analytics.
Incorta addresses these challenges by transforming the traditional approach to analytics with a unified platform that features direct data mapping—a precision engine that delivers performance, control, and visibility at a scale. By making a digital twin of an enterprise's data, mapping all the possible relationships, and developing a semantic/business layer, Incorta feeds inherent analytics-ready data to the downstream with context and efficiency at its core.
“Rather than having to build out elaborate schemas for staging and cleansing and then star schemas for dimensions and facts, our direct data mapping builds all that for you through smart data integration, providing all types of mappings and joins whether they’re defined in the database or not,” explained Balkenende. It gives “you full fidelity of the data [while] also being able to have the flexibility to get any data—no matter what the original question, KPI, or data that you needed…it will always be available within the Incorta platform for you.”
“I have yet to meet a company that isn’t trying to do more with the data that they have,” said Boris Jabes, co-founder of Census, a Fivetran company. However, this comes at a steep cost—literally. According to Gartner, bad data costs a company $13 million a year, and as Wakefield Research found, data engineers spend about 44% of their time building and maintaining data pipelines—which costs $520,000 per year.
To further exemplify the importance of unified analytics, Jabes introduced a Fivetran/Census customer—Canva—who sought to improve customer engagement with cross-channel insights. Canva soon realized that their challenge wasn’t data access—as any employee could simply access insights across a variety of channels—it was a lack of unification across these channels to provide a single source of truth.
According to Jabes, Canva saved $200,000 a year in engineering costs by centralizing their data with the Fivetran platform. Once centralized, Canva wanted to use that data to drive personalization at scale—which is where Census shined.
Helping to activate their data, Census easily turned Canva’s newly enriched data into automated decision making and operational efficiency.
“Census takes all the data sitting in your data warehouse…and can give you in-the-box, prebuilt pipelines that write the data into business applications, like Facebook Ads tools, like CRMs,” explained Jabes. “Census is really about the most efficient way to write data into those tools…you actually save money on those tools as well.”
Stephane Castellani, SVP marketing, CrateDB, added to the conversation by defining unified analytics as “the ability to work with all your data, from any source, in any format, and extract real-time insights in a consistent, governed way.” As a result, many organizations are rethinking their architecture, turning to architectures such as lakehouses, data mesh, and data fabric.
CrateDB, the distributed SQL database designed for real-time access to fast-moving data, regardless of structure and volume, centralizes diverse data to enable a variety of advanced use cases, including AI initiatives. CrateDB’s capabilities include:
- Handling multiple data types, including structured, time series, semi-structured, and even full-text
- Supporting real-time ingestion and querying without needing complex pipelines
- Auto-indexing and dynamic schema for simplified operations
- Cloud-native, distributed, and resilient, scaling automatically and staying
- available
- All SQL, enabling developers, data scientists, and analysts can work together without steep learning curves
CrateDB enables unified analytics, search, and AI in real time, capable of fitting into lakehouse, data mesh, and data fabric architectures. It unlocks use cases such as predictive maintenance, fleet tracking and logistics, customer-facing analytics dashboards, and feeding machine learning models with real-time features.
This is only a snippet of the full Achieving Unified Analytics: Top Tools and Strategies webinar. For the full webinar, featuring more detailed explanations, a Q&A, and more, you can view an archived version of the webinar here.