Data & Analytics for Growth Teams: Building Your First Data Stack

The Growth Engineer: A New Type of Role

Both Aurelien and Eliott have engineering backgrounds but work on business problems, not product features. Aurelien was one of the first engineers at Spendesk — but instead of building the product, he built the data infrastructure that powered the growth team. Eliott does similar work at Gorgias, a helpdesk for e-commerce.

"When you combine technical skills with a passion for business problems, you quickly end up in the growth ecosystem. The role is messy — your career path isn't well-defined, you don't know where you'll be in 6 months. But you get to create your own scope."

— Aurelien, Ex-Spendesk

This is the growth engineer archetype: someone who codes but thinks in funnels, someone who builds infrastructure but measures it in revenue impact.

Building the Data Stack: Three Phases

Phase 1: Foundations — Event Tracking

Before dashboards or models, you need events. Define the key user actions in your product and start tracking them consistently. Use a naming convention from day one (e.g., user_signed_up, campaign_created, payment_completed). Bad naming compounds into bad data.

Phase 2: Storage & Transformation — The Warehouse

Once you have events flowing, you need somewhere to store and transform them. A data warehouse (BigQuery, Snowflake, or even PostgreSQL for early-stage) becomes the single source of truth. Use dbt or SQL transforms to create clean, business-ready tables from raw events.

Phase 3: Activation — Dashboards & Automations

The output layer: dashboards for humans (Metabase, Looker, Amplitude) and automations for machines (reverse ETL to push cohorts into CRM, email tools, ad platforms).

What Makes a Great First Data Hire

"I was curious. I started coding at 12-13 because I wanted to understand how everything worked. With the internet, you just need a laptop and a connection — you can build anything. That curiosity is what makes a great first data hire."

— Aurelien, Ex-Spendesk

The ideal first data/growth engineer hire is:

• Full-stack capable: Can write Python, SQL, and basic frontend code.
• Business-curious: Wants to understand funnels, not just build APIs.
• Comfortable with ambiguity: The role is undefined — they need to create their own scope.
• Scrappy: Will build with spreadsheets and scripts before requesting a $50K/year tool.

Gorgias: Data at Scale

Eliott provides the perspective of a larger organization (150+ people when he joined). At Gorgias, there was already a team of engineers dedicated to go-to-market problems. His contribution was developing new capabilities around data activation — making sure the right information reaches the right person at the right time.

The challenge at scale: data quality. When 15 teams contribute data to the same warehouse, inconsistencies multiply. Ownership, naming conventions, and documentation become critical infrastructure.

Key Takeaways

1. Start with events, not tools. Define your key user actions before choosing any analytics platform.
2. Hire curious engineers. The best growth engineers are autodidacts who code AND think about business.
3. Name things well from day one. Bad event naming compounds into months of cleanup later.
4. Build for activation, not just reporting. The goal of data isn't dashboards — it's automated actions.
5. Data quality is infrastructure. At scale, invest in ownership, naming conventions, and documentation.

Based on a Growth.Talent LinkedIn Live session (63 minutes) hosted by Jeremy Goillot, featuring Aurelien (ex-Spendesk) and Eliott (Gorgias). Jeremy recruited Aurelien as the first business engineer at Spendesk — giving this discussion a unique insider perspective.