How Adidas Uses Databricks Genie to Power Data Analytics and Product Innovation

From Answering Questions to Asking Better Ones

Adidas shared a core insight about data analytics in an official post: truly great insights don't just answer existing questions — they help teams ask better ones. Simple as it sounds, this observation captures a broader shift in how enterprises approach data, moving from passive querying toward active exploration.

In traditional analytics workflows, analysts typically spend a significant portion of their time on data cleaning, wrangling, and computation, leaving far less time for actual business thinking. Adidas aims to change that by introducing tools like Databricks Genie — freeing analysts from tedious data processing so they can focus on uncovering deeper, more meaningful insights.

Databricks Genie's Role at Adidas

According to Adidas, Databricks Genie enables their internal analysts to spend less time crunching data and more time on the insights that actually matter. This shift in efficiency serves a larger purpose: helping teams make smarter decisions and design better products.

Databricks Genie is a natural language analytics capability built into the Databricks platform. Its core idea is to let users query data through conversation — without writing complex SQL every time. For a consumer goods company like Adidas, with massive data volumes and diverse business lines, lowering the barrier to data access is particularly valuable.

Data-Driven Product Innovation

Adidas emphasizes that its product innovation "runs on data." For a company built around athletic products, everything from material selection and design iteration to capturing consumer preferences depends heavily on making sense of large datasets. When the analytics process is accelerated and insights become more accessible, product teams can more quickly translate data into real design decisions.

This circles back to Adidas's opening point — good tools don't just help you find answers; they guide you toward questions you hadn't thought to ask. When analysts are no longer consumed by data processing, they gain the space to explore new directions that can genuinely drive innovation.

---

Databricks Genie falls under the category of "Text-to-SQL" or, more broadly, Natural Language Interface to Data (NL2Data). The underlying mechanism involves a large language model (LLM) translating a user's natural language question into structured queries (typically SQL), which are then executed against a data warehouse or data lakehouse and returned as results. Genie also introduces a "semantic layer" — pre-defining business metric definitions and table relationships — to reduce the likelihood of incorrect query generation and help the model accurately interpret questions from non-technical users. The key distinction from traditional BI tools is this: conventional BI requires analysts to build dashboards and fixed reports in advance, constraining users to a predefined view of the data. Conversational analytics, by contrast, allows users to ask open-ended, exploratory questions — where every query can theoretically be a novel combination. This makes it especially suited for early-stage exploration, when users don't yet know exactly what they want to ask.

---

Implications for Enterprise Data Strategy

The Adidas case, while brief, reflects several common priorities shaping enterprise data strategy today:

First, lowering the technical barrier to data analysis. Enabling business users and analysts to interact with data in more natural ways, reducing dependence on complex toolchains.

Second, directing human time toward higher-value work. Data processing can be automated, but business judgment and problem framing still require human thinking. The value of a tool lies in amplifying that capacity.

Third, ensuring data genuinely serves business outcomes. Whether the goal is smarter decisions or better products, the endpoint of data analytics is always business value — not data for its own sake.

It's worth noting that this content originated from a co-marketing effort between Adidas and Databricks, presenting results primarily through the lens of brand and platform promotion. Specific quantitative outcomes and technical details are limited. For organizations looking to draw lessons, it offers a useful framework for thinking — not a ready-to-replicate blueprint.

---

"Data democratization" has become a recurring theme in enterprise data strategy, referring to making data accessible and understandable to a broader range of people within an organization — not just data engineers or dedicated analysts. The path toward this goal has evolved through several phases: early efforts relied on self-service BI tools (like Tableau and Power BI) to lower the visualization barrier; a middle phase focused on data catalogs and governance frameworks to ensure data trustworthiness; and more recently, the maturation of LLMs has opened conversational querying as a new frontier. That said, data democratization is not without its challenges. When data quality is inconsistent, making data universally accessible can actually generate more erroneous conclusions. Natural language translation accuracy still has limitations in complex business contexts. And as access opens up, access controls and data security need to be strengthened in parallel. The Adidas case reflects an idealized vision — in practice, these friction points are very real and shouldn't be overlooked.

---

Takeaway

By leveraging Databricks Genie to streamline its analytics workflow, Adidas has freed analysts from data processing tasks so they can focus on generating insights and driving product innovation. It's a telling example of the broader trends toward data democratization and conversational analytics in the enterprise. And it leaves us with a reminder: the ultimate purpose of a tool isn't to answer questions faster — it's to help teams ask better questions.