DataTalks.Club Podcast Wiki

Wiki

Marketing to Analytics Engineering

Podcast-backed transition notes for marketers moving into analytics engineering through SQL, BI, dbt, product analytics, dashboards, and metric ownership.

Related Wiki Pages

Career Transition Analytics Engineering Analytics Engineering Roadmap Product Analytics dbt

Marketing to analytics engineering moves campaign and growth questions toward the modeled data behind those questions. The clearest archive example is Nikola Maksimovic moving from performance marketing into BI and analytics engineering at Ecosia. Her path starts with marketing reporting and Looker work. It then passes through SQL and BI projects. Data-pipeline literacy, dbt modeling, and product analytics support follow (marketing-to-analytics episode at 2:53-14:14).

The move builds on adjacent domain knowledge. Marketing gives the candidate familiarity with acquisition, conversion, audience splits, and stakeholder pressure. Analytics engineering adds reusable SQL models, tests, documentation, and BI-ready tables. It also adds shared metric definitions.

The transition works when marketing context becomes technical evidence. A dashboard becomes a governed model, and a campaign report becomes a metric definition. Funnel intuition becomes product analytics work (marketing-to-analytics episode at 18:34-39:36).

Start with these podcast discussions:

People connected to the cited interviews:

Adjacent pages:

Common Route

Marketing reporting creates the opening, and internal BI work creates the bridge. Analytics engineering becomes the target once the person can maintain reusable models. Maksimovic first valued performance marketing because it gave fast feedback on whether work was effective. When Ecosia moved to Looker, she helped build marketing-team reporting. She already understood the team’s numbers and questions (marketing-to-analytics episode at 2:53 and 7:18).

The transition then became explicit through conversations with the BI team. Maksimovic describes the first intended bridge as a marketing-analyst role. The required skills pointed toward BI work. SQL was the main gap, with data-pipeline understanding and Python basics behind it.

She also describes having to read and write more complex SQL. Larger data models mattered too, not just beginner queries (marketing-to-analytics episode at 8:45-12:50).

The final step is ownership of modeled analytical assets. Maksimovic’s later work included KPIs, dashboards, and product-team experiment support. It also included a dbt migration and wide-versus-narrow data-modeling tradeoffs. The stack included Looker, LookML, and Redshift. Airflow, Airbyte, and Snowplow also appear in the tooling discussion (marketing-to-analytics episode at 14:14-33:46).

That matches the broader analytics-engineering role described by Victoria Perez Mola. Her episode ties the job to SQL transformations, dbt tests, and documentation. DAGs and BI collaboration matter too (analytics-engineer skills episode).

Guest Differences

Guests differ on how formal the role boundary should be. Maksimovic’s team did not start with a clean analytics-engineer title. The work overlapped BI analyst, data analyst, and analytics engineer responsibilities, especially because the team was small (marketing-to-analytics episode at 25:06-28:40). Juan Manuel Perafan also warns that “between analyst and engineer” is an incomplete definition. His foundations episode emphasizes business reality, data modeling, software engineering practice, and safer analytical systems (foundations episode).

Guests also differ on tool centrality. Maksimovic says dbt strongly shaped the analytics-engineering title, but she still separates the role from the tool. Data modeling theory matters more than simply using dbt (marketing-to-analytics episode at 28:40-33:46). Natalie Kwong places dbt inside a broader ELT flow. Ingestion, warehouses, and orchestration keep the transition connected to the modern data stack (modern-data-stack episode).

The growth-oriented episodes add a second boundary difference. For Arpit Choudhury, the marketing-adjacent data path extends beyond dashboards. It includes event tracking, tracking plans, and warehouse transformations. It also includes BI, CDPs, and reverse ETL. That makes the marketing-to-analytics transition especially useful for people who want to specialize in data-led growth or data activation (data-led growth episode).

Transferable Marketing Context

The strongest transferable skill is knowing what a funnel means before writing the model. Maksimovic explicitly names comfort with marketing funnels, conversion funnels, and web acquisition funnels. She also names user journeys and touch points, plus the pressure to optimize and grow.

That background helped her support product managers and growth work. She could keep the user journey in view while building analytical outputs (marketing-to-analytics episode at 38:27-41:50).

This domain knowledge maps directly to product analytics. Maksimovic’s analytics-engineering work included growth analysis, retention analysis, RFM analysis, and NLP experiments. It also included dashboards and A/B testing support (marketing-to-analytics episode at 14:14 and 38:27). Choudhury’s growth-stack episode explains why those questions need event names, properties, source context, and ownership rules. They also need warehouses, BI, and activation tools, not only a campaign dashboard (data-led growth episode).

Marketing context alone won’t meet that bar. Perez Mola’s role episode centers analytics engineering on data modeling and quality checks. It also names pipelines, Looker, dbt, and Snowflake. SQL, DAGs, and collaboration matter too (analytics-engineer skills episode).

This transition therefore belongs with Data Analyst vs Analytics Engineer. Marketing experience helps with questions and definitions. Analytics engineering requires maintained models that other people can reuse.

Technical Learning Path

Maksimovic’s learning path puts Excel and pivot tables before SQL datasets. Real company queries help move the learner beyond beginner examples. She also recommends building dashboards in Looker or Tableau so the learner sees how modeled data becomes stakeholder-facing reporting (marketing-to-analytics episode at 41:50-45:09).

The next layer is data modeling. Maksimovic learned it through the dbt migration, where the work covered transformations and model organization. It also covered wide-versus-narrow tables and incrementalization tradeoffs (marketing-to-analytics episode at 18:34-33:46). That aligns with the Analytics Engineering Roadmap, which moves from analytical SQL to reusable models and tests. Documentation, metric ownership, and domain specialization follow.

Pipeline literacy is a real requirement, but the marketing path in this archive is SQL-first rather than Python-first. Maksimovic names SQL as the main skill, then pipeline understanding and Python basics as supporting gaps (marketing-to-analytics episode at 9:53-12:50). Kwong’s modern-stack episode adds the surrounding system. That system includes ingestion, warehouse storage, dbt transformations, and orchestration. It also includes CDC and reverse flows (modern-data-stack episode).

Portfolio and Internal Proof

The archive favors internal proof when it’s available. Maksimovic didn’t jump directly from marketing into an external analytics-engineering role. She used marketing reporting, a Looker migration, BI-team conversations, and BI projects alongside marketing work to prove the transition (marketing-to-analytics episode at 7:18-14:14 and 23:12).

For an external portfolio, the same evidence should be made visible. A strong marketing-to-analytics project should define campaign or product events. It should model funnel or retention tables, document metric grain, add dbt-style tests, and publish a dashboard from shared models.

That standard comes from Maksimovic’s dbt, Looker, product analytics, and modeling work. It also comes from Perez Mola’s emphasis on dbt tests and documentation (marketing-to-analytics episode at 18:34-33:46, analytics-engineer skills episode).

Good project themes for this transition include these options:

These fit the podcast evidence because Maksimovic covers marketing funnels and product support. Her episode also covers A/B testing and RFM analysis. dbt modeling and Looker appear there too. Choudhury connects modeled growth data to activation workflows (marketing-to-analytics episode at 38:27-41:50, data-led growth episode, Analytics Engineering Portfolio Projects).

Sponsorship and Team Structure

The transition is easier when the existing organization has a BI or data team that can sponsor practical work. Maksimovic describes conversations with a BI colleague and a BI analyst who helped her learn Looker. She also describes later teammates who served as mentors once she joined the BI team (marketing-to-analytics episode at 9:53 and 45:09-50:23). That makes this path a strong fit for marketers in companies with existing reporting needs. Dashboard migration, dbt adoption, and product analytics needs also create openings.

Small-team environments can blur the role boundary. Maksimovic’s BI team was small enough that people did both analysis and analytics-engineering work (marketing-to-analytics episode at 25:06-28:40). Tammy Liang gives a related team growth example. Dashboards and business-health monitoring grew into a warehouse, dbt, Data Studio, and Notion documentation. Testing, monitoring, forecasting, and adoption work followed (building-and-scaling-data-team episode).

Use these pages for the role, stack, and adjacent transition context.