Contract Python Data Analysis Jobs in Vancouver, BC

Jetson has gone from $20M to $100M in run-rate sales this year. We're on track to end the year around $250M and scaling toward $1B. That growth is being driven by a marketing org that is getting more sophisticated every quarter - performance marketing across Meta, Google, direct mail, OOH, and CTV; a strong in-house creative studio; an emerging partnerships function; and an experimentation and lifecycle team that needs sharper attribution and faster feedback loops to keep compounding.

The data layer underneath all of that is what unlocks the next leg of growth. We've built out a real data lake - Iceberg on AWS Glue, dbt for transformation, Fivetran for ingestion, Metabase for BI. The core framework exists. Meaningful work remains to refine it, fill the gaps, integrate the next wave of sources, and ensure the team has accurate real-time data they can trust.

You'll own the marketing data lake end to end - the integrations that bring data in, the bronze/silver/gold models that shape it, the dashboards that surface it, and the analysis that turns it into decisions the team actually makes.

This is a full-stack role. On any given week you might be standing up a new Fivetran connector for a direct mail vendor, writing the dbt models to land it cleanly into our medallion architecture, building the Metabase dashboard for the performance team to use, and then pulling the analysis together to tell them whether the campaign worked. There's no handoff between the engineering and the analysis - you do both.

You'll work closely with the performance marketing team (who'll lean on you for measurement and incrementality), the growth and experimentation team (who'll need experiment infrastructure and LTV models), and our core data engineering team (who own the broader platform you're building on top of).

The marketing data lake. Iceberg, Glue, dbt, Fivetran, Metabase. You own the bronze/silver/gold models, the schemas, the freshness, the cost, and the contract with downstream consumers. If a number is wrong in a dashboard, you're the one who finds out why and fixes it at the source.

Source integrations. Every new channel, vendor, or partnership comes with a new data source - Meta and Google Ads, HubSpot, direct mail vendors, CTV platforms, call tracking, utility and financing partners. You stand up the Fivetran connectors (or write the Python when Fivetran can't), land the raw data in bronze, model it through silver and gold, and surface it where it needs to be.

Attribution and measurement. Multi-touch attribution across paid, organic, direct mail, OOH, CTV, partnerships, referrals, and field sales. Identity resolution across web, CRM, and offline conversions. You'll design and maintain the attribution models the team relies on day to day, and the incrementality work that informs bigger budget decisions.

Experimentation infrastructure and readout. Geo holdouts, conversion lift studies, direct mail incrementality, on-site experiments. You'll build the data infrastructure that makes them trustworthy - assignment logging, variant tracking, exposure data - and you'll do the analysis to tell the team whether the result is real.

LTV, conversion, and channel ROI. The models that tell us what a customer is worth, what each channel is actually delivering, and where the next dollar should go.

Dashboards and self-serve. The Metabase dashboards leadership, marketing, and sales check daily. Built so the team can answer their own questions when they can, and knows when to come find you when they can't.

Being the analytics partner to the GTM team. The marketing, performance, and exec teams will bring you real questions all the time - whether a campaign worked, why one market is lagging another, where we should be investing more. You answer those questions with data, clearly, and with enough context that the team trusts the answer.

• 5+ years working on marketing, growth, or revenue data. Pure backend or pure analytics engineering backgrounds aren't quite right - we need someone who's been close enough to the marketing problem to know where the hard parts are.
• Deep dbt fluency. You've shipped real production dbt projects, understand medallion architecture, and have opinions on testing, contracts, and model design. dbt is non-negotiable.
• Hands-on with a modern data lake or warehouse stack. Iceberg, Glue, Fivetran, and Metabase are our tools, but we care about the patterns more than the specific vendors. Equivalent work on Snowflake/BigQuery/Databricks with comparable tooling translates directly.
• Strong SQL and Python. Deep SQL - you think in SQL, not around it. Python for the integrations Fivetran can't handle, for the stats SQL can't do cleanly, and for everything in between.
• You've built attribution before, not just consumed it. You understand the difference between last-click, MTA, MMM, and incrementality testing, and you know why each one is useful and where each one breaks.
• You've supported experimentation at scale. Geo tests, holdouts, conversion lift, incrementality - pick at least one and have a real story about it.
• You build dashboards that get used. You make thoughtful design decisions, not just drop fields onto a canvas. You know the difference between a dashboard that answers questions and one that creates them.
• You care about data quality the way good software engineers care about test coverage. Tests, contracts, monitoring, lineage. You don't ship a model without thinking about how you'll know when it breaks.
• AI-native. You've used Claude Code, Cursor, or similar tools to ship real work. You think about how AI changes the leverage of a one-person data team, not whether it does.
• You communicate clearly with non-technical marketers. Most of your stakeholders won't speak SQL. You can explain why a number changed without making them feel small for asking.

• Direct experience with our stack - Iceberg, AWS Glue, Fivetran, dbt, Metabase.
• Experience with marketing-specific sources: Meta and Google Ads APIs, HubSpot, call tracking platforms, direct mail vendor data, CTV reporting.
• Experience at a company that scaled from low nine figures to ten figures in revenue.
• Experience with identity resolution and customer data platforms.

This is the highest-leverage data hire we'll make this year. The decisions made off the data you build will move millions of dollars in ad spend, shape channel mix, and define how we measure the business as we scale to $1B. The infrastructure work is real - there's a clean greenfield problem here that few companies our size get to solve properly. And you'll be doing it with a marketing org that has the budget, the ambition, and the technical literacy to use what you build.

Apply at [link]. There's a short questionnaire as part of the application - we read every one.

If you've built marketing data infrastructure that you're proud of, tell us about it. Specifics matter more than scope.