You will build a company's entire marketing data layer from scratch. Not maintain someone else's, not tweak dashboards. Build it, own it, and decide what it runs on.
Right now there are two disconnected warehouses and no way to answer a simple question like what ROAS looked like this week across the company. That is what you are here to fix.
US company, fully remote team across Europe, B2B contract.
Rate: $5,000 to $7,500 per month, up to $10,000 for someone exceptional. Including VAT.
Some companies call this role Analytics Engineer or Growth Data Engineer. Here it is 90 percent engineering.
What you'll do
• Build ingestion pipelines in Python against the Google Ads, Meta, Apple Search Ads, AppsFlyer and RevenueCat APIs, with the rate limits, schema changes and backfills that come with them
• Design the data model and warehouse from scratch. Postgres today, but you pick what actually holds up at this volume
• Write the identity resolution layer yourself. Deterministic and probabilistic matching across web and mobile, click IDs, device identifiers, deduplicated event streams
• Solve performance at click-level scale. Partitioning, indexing, query planning, storage cost. This will be the heaviest load in the entire stack
• Merge two existing warehouses into one model, including migrating the legacy Python pipelines or replacing them entirely
• Build the reconciliation logic that catches when platform numbers and your own numbers disagree, then find out why
What this role is not
Not a BI or dashboarding role. Not an analytics role where SQL and a modelling tool cover the job. Not a marketing role with a technical flavour. And not a lead position, at least not yet. You are the first person doing this here.
What you bring
• 5+ years as a data engineer, with real ownership of a marketing or attribution data layer you built yourself
• Strong Python, production pipeline code, not scripts
• Deep understanding of database performance at high volume
• Hands-on experience with at least one ad network, Google Ads or Meta preferred, including the platform itself
• Proven work on attribution and user deduplication across web and app
• Background in consumer mobile apps, gaming or e-commerce, where volume and multi-channel complexity are the norm
• Able to explain technical decisions to a non-technical VP of Growth, and turn vague requests into something that works. This is where most candidates so far have fallen down
You don't have to tick every box. If you have owned this kind of system end to end, the rest follows.
Nice to have
• Google Ads and Meta both
• An MMP such as AppsFlyer, Adjust, Singular or Branch
• Cross-platform subscription logic, someone subscribing on web and later in the app
• SKAdNetwork, consent handling, click-ID persistence, server-side events
• Migrating Postgres to an analytical warehouse
• Modelling ROAS, LTV, CAC and payback
Knock-out criteria
• 4 years of experience or less
• Repeated short stints across companies
• No exposure to any MMP or attribution tooling
• Claiming 100 percent attribution accuracy, which is not possible
• Not available around the daily stand-up at 15:00 UK time
Practical
Remote anywhere in Europe. Daily stand-up at 15:00 UK time, available an hour either side. Beyond that you set your own hours.
B2B contract, monthly fee, output-based, no time tracking. Three month ramp-up period. Roughly 14 days off per year, not tracked.
Why this one
Full stack freedom. DigitalOcean and Postgres today, but nothing is sacred. BigQuery, ClickHouse, Snowflake, whatever you can defend. You can rip out the existing Python if you want to.
First hire in this area, so you set the architecture instead of inheriting one.
Interviews start next week. I run this search directly with the hiring team, so you get real feedback fast instead of silence.
Tags: Data Engineering, Python, Attribution, Marketing Analytics, AppsFlyer, Google Ads, PostgreSQL, BigQuery, ETL, Growth