Analytics Engineering: What the Role Is, What It Pays, and How to Get In

Analytics engineering emerged as a distinct role around 2016–2019, driven by the convergence of SQL-first transformation tools (dbt), affordable cloud warehouses, and the recognition that the gap between raw data and usable analytics needed dedicated engineering attention. Today it is one of the fastest-growing roles in the data profession. Here is what the role actually entails, what skills it requires, and how careers develop within it.

What Analytics Engineers Actually Do

The analytics engineer role sits at the boundary between data engineering and data analysis. Data engineers build pipelines that move and load data; data analysts build reports and answer business questions. Analytics engineers build the transformation layer in between — the clean, tested, documented data models that analysts use for analysis and BI tools consume for dashboards.

In practice, the analytics engineer's work includes:

**Transformation and modelling:** Writing dbt SQL models that clean raw ingested data, implement business logic, and produce the dimensional model (fact tables, dimension tables, mart tables) that downstream consumers depend on. This includes staging models that standardise source data, intermediate models that join and enrich entities, and mart models optimised for specific analytical use cases.

**Data quality and testing:** Writing dbt tests that enforce data quality constraints — uniqueness, not-null, referential integrity, business logic assertions. Running those tests in CI/CD pipelines. Investigating failures and tracing them to root causes.

**Documentation:** Writing field descriptions, metric definitions, and model-level documentation in dbt's schema.yml. The analytics engineer is the person who makes the transformation layer legible to analysts who did not build it.

**Collaboration with analysts:** Analytics engineers work closely with business analysts and data analysts to understand what data they need, what metrics they use, and how the data model can be designed to serve their use cases. The analytics engineer translates business requirements into data model design.

**Collaboration with data engineers:** Analytics engineers work with data engineers on the ingestion layer — what sources to connect, at what granularity, with what refresh frequency. They may build or maintain source connectors in addition to transformations.

**Performance and cost management:** Monitoring query costs in the warehouse, identifying expensive models, optimising slow transformations, and managing incremental materialisation for large tables.

The Skills Required

### SQL

SQL proficiency is the baseline. Analytics engineers write SQL constantly — modelling, transformation, data quality investigation, ad-hoc analysis. The level required is beyond basic SELECT/WHERE/GROUP BY: window functions, CTEs, complex joins, subqueries, and warehouse-specific SQL features (Snowflake's QUALIFY, BigQuery's STRUCT/ARRAY, Redshift's distribution and sort keys).

The analytical SQL skills — using window functions for cohort analysis, running totals, period comparisons — are what differentiate senior analytics engineers from junior ones. The ability to read an explain plan, understand query cost, and optimise slow queries for a specific warehouse architecture is expected at senior levels.

### dbt

dbt is the dominant analytics engineering tool and the one skill closest to a universal requirement for the role. Hiring managers expect: dbt model materialisation (table, view, incremental, snapshot), Jinja templating and macros, testing with schema.yml and singular tests, documentation practices, and deployment via dbt Cloud or CI/CD pipeline.

At senior levels: dbt packages, complex macro development, slim CI implementation, incremental model architecture including late-arriving data handling, and dbt project structure best practices.

### Python (for some roles)

The Python requirement varies by team. Some analytics engineering roles are SQL-only; others require Python for custom data transformations, pipeline scripting, or working with source APIs. At minimum, enough Python to work with notebooks for exploratory analysis and to write simple data manipulation scripts is useful. For teams using Databricks or Spark, Python proficiency is more significant.

### Version Control and CI/CD

Git is non-negotiable. Analytics engineering code lives in git; PR-based workflows and code review are standard practice. Understanding branching, merge workflows, and using git in a team context is a baseline skill.

CI/CD for data pipelines — configuring GitHub Actions or similar to run dbt tests automatically on PR — is expected at mid-to-senior levels.

### Data Warehousing Concepts

Dimensional modelling (fact tables, dimension tables, star schema, slowly changing dimensions), OLAP vs OLTP patterns, and how columnar storage and query optimisation work in cloud warehouses are foundational knowledge. An analytics engineer who does not understand why distribution keys matter in Redshift, or why partition pruning affects BigQuery costs, will make systematic architecture mistakes.

### Business Domain Literacy

Analytics engineering is not a purely technical role. The analytics engineer needs to understand the business domain well enough to make good modelling decisions — what constitutes a customer, how revenue recognition works, what an "active user" means for this specific product. Domain literacy comes from time spent in the business and from close collaboration with analysts and business stakeholders.

The Role in Different Organisations

### Early-Stage Startups (Pre-Series B)

Analytics engineering at an early-stage startup is often a generalist data role — the analytics engineer may also be the data engineer, the data analyst, and the data infrastructure manager. The stack is simpler (typically a single cloud warehouse plus dbt plus one BI tool), but the breadth required is significant.

This is actually an excellent environment for building foundational analytics engineering skills quickly: you make decisions about everything from ingestion to visualisation, you see the consequences of those decisions directly, and the feedback loops are fast.

### Mid-Market Companies (Series B to enterprise)

The specialised analytics engineer role is most clearly defined in mid-market companies. There is a defined data engineering team handling ingestion and infrastructure, defined business analysts or BI developers handling dashboard and report production, and the analytics engineer building and maintaining the transformation layer between them.

### Enterprise

Large enterprises often have multiple analytics engineering teams aligned to business domains — a commercial analytics engineer, a product analytics engineer, a finance analytics engineer. Specialisation is deeper; the domain knowledge requirement is higher; the cross-functional collaboration is more complex. Governance and standards become more important at this scale.

Compensation

Analytics engineering compensation varies significantly by geography, company stage, and seniority. Rough ranges in the United States (2025):

- **Entry level (0–2 years):** $90,000–$130,000 total compensation

- **Mid-level (2–5 years):** $130,000–$180,000 total compensation

- **Senior (5+ years):** $180,000–$250,000+ total compensation

- **Staff/Principal:** $220,000–$300,000+ at larger companies

In London, ranges are roughly 60–70% of US levels. In Sydney and other major Australian cities, roughly AUD$90,000–$180,000 depending on seniority.

High-growth technology companies (FAANG, fintech, SaaS at scale) pay at the upper end; traditional enterprises pay at the lower end for equivalent seniority.

Career Paths Into Analytics Engineering

**From data analysis:** The most common path. Data analysts who learn dbt and develop SQL depth move naturally into analytics engineering. The business domain knowledge carries over; the new skills are the engineering practices (version control, testing, CI/CD) and more advanced SQL.

**From software engineering:** Software engineers who develop interest in data — often through working on data infrastructure problems — move into analytics engineering. The engineering practices (git, testing, CI/CD) are already in place; the new skills are SQL depth, dimensional modelling, and business domain literacy.

**From business intelligence development:** BI developers who spend time in Tableau, Power BI, or Looker and develop frustration with the data quality and model design problems they encounter in their source data often move into analytics engineering to address those problems upstream.

**From data engineering:** Data engineers who find the infrastructure work less engaging than the analytical work, and want to be closer to business users and analytical outcomes, move into analytics engineering.

The Career Ceiling Question

Analytics engineering is a relatively new specialisation, and the senior career path is still developing. The question of "what does a staff analytics engineer look like?" does not have a fully settled answer at most organisations.

The paths that experienced analytics engineers take: senior/staff analytics engineer (technical depth, project leadership, standards ownership), analytics engineering manager (team leadership, hiring, programme management), head of data or data team lead (broader scope including strategy), and data architecture (moving into platform and systems design).

The role rewards technical depth, business domain knowledge, and the ability to operate at the intersection of engineering rigour and analytical need. Those three skills together are rare and valuable.

Our data engineering consulting practice includes analytics engineering delivery and teams that embed with client data organisations — contact us if you are building or scaling an analytics engineering capability.

A former Microsoft data architect audits your data foundation, identifies your top priorities, and sends you a written plan. Free. No pitch.