Loading post…
DatabricksElizabeth Dobbs, Thomas Russell, Katy Yuan, Sydney SundellProblem & solution
How Databricks’ marketers use data 3x more with Genie, an AI analytics assistant
Summary
Databricks marketers struggled to get trusted answers from fragmented marketing data quickly, prompting the creation of Marge, a marketing implementation of Genie Agents. The team first unified campaign, sales, web, event and advertising information in a governed Marketing Lakehouse, then grounded the assistant in business definitions, verified logic and example queries. Marge translates natural-language questions into analytical queries, uses Unity Catalog for lineage and role-based access, and can ask clarifying questions or perform multi-step analysis. The reported results include threefold greater data use in decisions, more than 85% adoption, 50% quarter-over-quarter growth, over 5,000 answered questions and a 25% reduction in flagged incorrect responses. Databricks attributes the rollout to continuous feedback, benchmark evaluation, focused use cases and embedding Genie in existing workflows rather than treating conversational analytics as a one-time deployment.
Context
Marketing data was distributed across campaign platforms, web analytics, CRM systems, event tools, advertising channels and sales data, with inconsistent definitions and reporting logic. Databricks created a governed Marketing Lakehouse shared across Marketing, Finance, Sales and Product to provide consistent go-to-market metrics and support self-service analytics.
Approach / What changed
Databricks built Marge with Genie Agents and grounded it in Marketing Lakehouse data, business terminology, documented relationships, verified metric logic, example SQL and trusted assets. Unity Catalog supplies metadata, lineage and role-based access controls. The team improves the assistant through user feedback, monitoring, benchmark questions and focused domain agents, while rolling it out from one bounded use case to broader workflows.
Takeaways
Marge translates natural-language marketing questions into analytical queries over governed enterprise data, while Unity Catalog limits answers according to each user’s authorized access.
Accuracy work combines documented table relationships and business definitions with verified logic, example SQL, trusted assets, clarification instructions and recurring benchmark evaluation.
Databricks began with a bounded, frequent use case and expanded through user demand; one BI manager currently spends about one hour per week reviewing feedback and maintaining Marge.
FAQ
What was the best first use case for Databricks’ marketing Genie?
Databricks recommends a frequent, well-defined question supported by a small set of trusted data. Its initial example was email campaign performance using campaign, recipient and engagement tables, followed by expansion based on user demand.
How does Marge improve the accuracy and trustworthiness of its answers?
The team documents data relationships, metrics and business terminology, then adds verified answers, example SQL and trusted assets for important questions. Users provide feedback, while the analytics team reviews issues and tests benchmark questions to identify regressions.
Does Genie replace marketing analysts?
Genie handles many repetitive and lower-complexity questions, allowing analysts to spend more time on experimentation, model design and strategic analysis. Databricks says this broadened marketing data usage without adding analytics headcount.
Source and freshness
- Source
- databricks.com
- Published
- Added to Yomu
Summaries are generated from the original by a language model. Check the original before relying on it.
Keep reading
Related reading
The Good Stuff
Paid picks we use ourselves. Links carry a referral code — a subscription through one earns us a commission at no extra cost to you.
The best engineering reads, once a week
One useful email. Unsubscribe anytime.