The semantic layer concept within the data stack is not new but is an increasingly popular topic of conversation. I predict that in 2022, we’ll see mainstream awareness of the semantic layer, especially as enterprises begin to see real-world examples of its benefits.
The fact that industry leaders are discussing the need for a semantic layer won’t hurt either. Gartner recently published a definition of the semantic layer, calling it “a business representation of data that helps end users access data autonomously, using common business terms. The semantic layer achieves this by mapping complex data into familiar business terms such as product, customer or revenue to offer a unified, consolidated view of data across the organization.”
In addition, through recent product news, Google itself endorsed the semantic layer as the key enabler of governed self-service analytics, and the best way to scale artificial intelligence (AI) and business intelligence (BI) programs in the enterprise. I couldn’t agree more.
In addition to the increased awareness, why will we hear more about the semantic layer in the coming months, you may ask?
Well, over the course of the past few years, cloud data lakes and cloud data warehouses have become well-accepted data platform architectures. In addition, most enterprises currently operate cloud data platforms or plan to deploy them soon. This means that more data can be collected and stored than ever before, since the data platform’s scaling and management can be outsourced to expert cloud partners.
Contrary to many predictions, BI tools (including mainstay Excel) have continued to proliferate, and the data scientist role has emerged as another data-hungry consumer, needing the same access to business-friendly data as their business analyst partners. Because of this, the analytics landscape has become even more daunting for IT and users alike. Put simply, there is more data (in volume and variety) and more consumers wanting to use said data.
The bottom line is that businesses will have to define a semantic layer eventually. If you don’t create one for your data users, then they will do it for themselves, using whatever technology they believe is easiest. If this happens, commonality and correctness will become more difficult and will hurt data use, while increasing time to analysis. Simple and secure access to clean, understandable data is needed by all within an organization, be they a data scientist, an analytics expert, or a BI user. There is a real need to present common access to your company’s data, as the alternative perpetuates a messy situation that becomes harder and harder to rebound from.
Gartner has also noted the various approaches to implementing a semantic layer in the data stack, including within the BI platform, within a data warehouse, within the data integration layer, or within a virtualized access layer. Technology solutions in all these categories are starting to adopt the term “semantic layer” to describe their capabilities. This can only lead to greater awareness – but with this increased awareness, there needs to be an increased understanding of the right way to think about a semantic layer.
The key is commonality. By defining their business metrics, data access, and data transformations in a single location, companies can then guarantee that the whole organization is speaking the same language, regardless of their job roles, use cases, or the tool sets they’re using. The adoption of a semantic layer will simplify the data stack, and ensure all users are playing by the same rules. Companies will find that Data Governance is easier and that as an added benefit, they’ll have freed themselves from vendors’ proprietary chains, creating the flexibility they’ll need as new data platforms and tools inevitably proliferate in the coming years.
As we head into the new year, the underlying effect of industry experts adopting the terminology and the growing need for data access will lead enterprises to consider a semantic layer as they evaluate their broader data and analytics strategy.