How Omni Analytics’ flexible, front-end focused approach redefines the traditional BI model by promoting business-driven analytics. We compare Omni’s methodology to Looker’s structured LookML, highlighting how creating safe spaces for ephemeral analytics can improve model usability, performance, and long-term success.
Key Takeaways:
As someone who spends a lot of time working on BI implementations and migrations, I’ve had the chance to see how different tools and methodologies play out in the real world. Two approaches that often come up in conversations are the traditional LookML model used by Looker and the more flexible, front-end focused approach taken by Omni Analytics. Both have their merits, but the differences between them can significantly impact the long-term success and maintainability of your BI environment.
Looker’s LookML is well-known for its structured, opinionated approach to building a BI model. It’s designed to create a common set of metrics across an organization, ensuring consistency and accuracy in reporting. The model lays down clear paths for how data should be understood and used, which is particularly valuable in large organizations where maintaining a single source of truth is critical.
However, as use cases evolve and become more specific, the requirement to push the vast majority of business logic into LookML becomes counter productive. Teams often create one-off models or custom metrics to handle specific needs that don’t fit neatly into the existing framework. Over time, this can lead to a model that’s cluttered and difficult to maintain, undermining the very consistency and clarity it was meant to provide.
Omni Analytics takes a different route, one that might be more in tune with the natural flow of how businesses evolve. Rather than enforcing a strict model from the start, Omni empowers users to build analytics on the front end, which can then be promoted into a shared semantic layer once there is consensus. Their approach is akin to creating desire paths—those natural trails that form as people find the most efficient routes through a landscape.
By creating safe spaces for ephemeral analytics, you will avoid compromising the overall integrity of the system because the essential metrics and structures remain intact. The result is a BI environment that stays organized and maintainable over time, even as the business’s needs evolve.
Two things to consider: First, if your business is growing or rapidly changing, it may be beneficial to let business requirements drive the logic that gets shared rather than starting with an opinionated model. This approach allows for faster data iteration and results in a cleaner, more approachable model—a key virtue of a strong self-service BI experience.
Second, many organizations are currently considering migrations due to cost, support issues, or the desire to partner with a company that is more focused on innovation. For those contemplating a migration from Looker or other solutions, we view Omni's approach not as a trade-off but as a significant innovation. It maintains what made Looker great while offering a better overall experience.
