Data vendors face the challenge of standing out in a crowded market. Many data products appear undifferentiated, with unclear value propositions. Successful data companies distinguish themselves by offering innovative solutions that are easy to implement without diverting significant resources.
Although powerful, traditional BI tools like Tableau or ClickView can lead to inconsistencies in KPI definitions when each dashboard independently defines these metrics. This issue can result in different, albeit technically correct, numbers being reported for the same KPI, causing confusion and reducing trust in the data.
While a semantic layer helps maintain standard definitions across dashboards, users can still tweak metrics and create their own versions of the truth. However, a standardized layer provides a common ground, allowing users to benchmark and validate their custom dashboards against a trusted source.
The adoption of semantic layers faces challenges due to incomplete solutions, wherein existing tools only solve part of the problem. Additionally, there's a disconnect between those implementing semantic layers and those benefiting from them, leading to potential reluctance to adoption.
The future of semantic layers lies in decoupling them from specific BI tools and integrating them more seamlessly into data warehouses. This would enhance interoperability and scalability, potentially increasing adoption across organizations, particularly those transitioning to a more data-driven approach.
"Unfortunately, data vendors in the market all come across as quite undifferentiated. Their value proposition does not always tend to be clear. They all make promises that are very cliché."
Alessandro Pregnolato
"Observability has been very interesting as an emerging category, and it's been emerging for a few years, but it's still something that lots of talk and focus on."
Alessandro Pregnolato
"Now that we have faster data warehousing, the semantic layer is finally enabled to bring all its value."
Alessandro Pregnolato