We recently worked with a client that had both Power BI and Tableau in production at the same time.
Power BI was heavily connected to the customer’s Microsoft ecosystem and Tableau was covering
older, more complex reporting. This multi-tool ecosystem didn’t happen overnight and came by way
of decades of growth and picking the right tools for the job. This client had similarly picked Databricks
to be its enterprise data platform for the same reason: it is the right tool for the job.
For computational power, governance, and AI, Databricks has always been a leader, but the release
of AI/BI Dashboards and Genie has opened a new chapter for this client. Importantly, there was never
a discussion of “why should we procure this new BI tool?” Procurement had already been completed –
there were no new licenses, no additional fees. On top of this, we had already created a sophisticated
medallion architecture for massive amounts of data over the preceding years, so the key data was in
place and all that was left to do was explore.
What Triggered the Evaluation
That ecosystem of decades of right decisions is extremely challenging to hold together in all too familiar
ways. The client’s PowerBI platform required its own management team and a substantial access request
framework. Tableau was hosted on virtual machines with a second management team, vulnerability
management, and increasing costs year-over-year. Worst of all, both platforms have inherent license
limitations that required the client to restrict usage to certain functional areas.
Headcount grew by orders of magnitude, but the number of Creator/Pro seats stuck in the double
digits. Onboarding new users brought pained questions of “do you really need to create?”
Because Databricks was already in place and both of the traditional BI platforms had created pain points
for the organization, we decided to explore how AI/BI Dashboards and Genie could fit in the larger
ecosystem and if any functionality could be migrated.
Where AI/BI Dashboards and Genie Moved the Needle
A key mission for this client is data sharing across teams for better organizational understanding. Visual
mediums are already great for communication, but only certain technical users can create effective
reporting with Tableau and Power BI. Additionally, change requests can burn weeks of back-and-forth
iteration trying to get to underlying asks. AI/BI Dashboards and Genie were fundamentally different for
this client because the change request data discovery pattern became partially irrelevant! Conversational
analytics with Genie allow you to go beyond what is being presented and understand where it’s coming
from with full contextualization:
- “Why did consumption spike in May?”
- “What does ‘burn rate’ mean here?”
Democratization is also achieved through allowing all workspace users to create with their data. The
licensing posture of the traditional BI tools made it so that reporting was a bottleneck. We have seen
hundreds of dashboards get created with AI/BI Dashboards in the past four months because those
restrictions are not in place. Unfamiliar and even non-technical users have leveraged the conversational
experience to get on-task training while building their own dashboards, functionally increasing the
number of developers for the client. This directly maps to organizational excellence with the data sharing
mission because insights are crystallized at the source instead of translated downstream.
AI/BI Dashboards and Genie brought more of the organization in, both for understanding and creating. In
this way, it has been a completely different experience from traditional BI.
The Path Forward
Although AI/BI Dashboards is not yet at full feature parity with Tableau or Power BI, it is serving a wider
purpose than traditional BI for our clients. Having your reporting directly on top of your governed data,
with access to a single pane of glass for your enterprise data and semantic model brings simplicity and
peace of mind. On top of this, the differentiating democratization of being able to chat directly with your
data to get beyond static reports to go to the why opens up discovery to all of your team. For these
reasons, the client has committed to phasing out Tableau in favor of AI/BI Dashboards and Genie to
reduce the total cost of ownership and increase speed from ingestion to insights. Power BI is still in
place because of the richness of the underlying Microsoft ecosystem for unstructured and
semi-structured sources, but we will continue to evaluate as feature parity is fully achieved.
In short, AI/BI Dashboards and Genie make the most sense when a customer is already using Databricks
and getting value from it as an enterprise data platform. There are no new licenses or ingestion hurdles
to clear, and the potential to have a 24/7 data analyst to support insights and development is too hard to
ignore. Similarly, AI/BI Dashboards alone, without a mature Databricks platform underneath, provide
limited value (even with Genie) compared to traditional BI because the underlying simplicity around data
interactions goes away. For existing Databricks customers, we recommend a trial to explore the strangler
pattern for high-value dashboards in parallel with art of the possible conversational analytics to bring in
new users. AI/BI Dashboards and Genie are already able to displace these traditional BI powers in some
contexts, and we expect this ability to expand going forward.
