Attribution that fits your acquisition
Propel designs and builds bespoke measurement systems for marketing teams.
Your attribution is in conflict with the truth.
It is telling you to pull budget from brand and push it into paid search. It is telling you 50% of your traffic is direct. It is attributing your best customers to the last click before purchase.
You cannot prove it wrong. And you cannot change it.
The problem is not that attribution is broken. It is that it was built for someone else's business — by a tool vendor who does not know your channel mix, or a data team without marketing context.
Attribution projects underdeliver because the people who build them don't understand marketing.And the people who run marketing can't interrogate what gets built.
Between us founders, Barbara and Timo have spent decades on either side of that gap. We built Propel to sit exactly at the seam.
Barbara — Started her tech career as a growth lead in different startups. She then spent nearly 10 years as an independent consultant, helping companies improve marketing performance with data initiatives. She writes about marketing measurement and analytics at 021 Newsletter.
Timo — Built data infrastructure for marketing teams and kept hitting the same wall: technically correct systems the marketing team couldn't use. That frustration is what led him to start working at the intersection instead.
We start with the marketing, then work backwards.
What campaigns are you running? What do you want to run but cannot measure yet? The answers to those questions shape everything we build. Models, data products, pipelines — all of it is reverse-engineered from your marketing reality, so it stays useful to the marketing team from day one.
In the end, our goal is to enable both sides of the table: marketing and data teams.
Good attribution is not just a report. It is infrastructure that unlocks things that were not possible before.
For marketing teams
Know what is actually driving growth versus what was going to happen anyway.
Linear TV and out-of-home become measurable through geo proxies and signal engineering.
Replace gut-feel budget defense with evidence, even for channels that do not click.
Reliable value signals let platforms bid on what actually matters, not just who converts cheapest.
For data teams
Owned attribution infrastructure you control, at a fraction of the ongoing cost.
Attribution signals structured for automated analysis and AI-powered reporting.
Use attribution as a foundation for LTV prediction, churn modelling, and acquisition forecasting.
One attribution model the whole organisation can query, interrogate, and build on.
We cover the three areas where attribution most commonly breaks.
The attribution model your team can actually interrogate. The measurement foundation it depends on. And the advanced techniques that only become possible when both are right.
Triangulated, transparent attribution
Replace a black box with logic your team can see, interrogate and build on.
Most teams rely on a single model they didn't choose and can't interrogate. We replace that with explicit logic showing multiple perspectives, so budget decisions come with real confidence.
This often means going beyond standard inputs. Things like:
- Survey responses
- Voucher redemptions
- Vanity URLs
- Geo traffic patterns
- MMM and incrementality outputs
These can reflect how your business actually grows better than any pixel. Everything we build is documented and owned by your team.
A single model nobody chose, outputs nobody fully trusts. Assumptions invisible. Budget conversations go in circles.
Explicit, multi-perspective attribution. Survey, voucher, geo and MMM signals all triangulating. Assumptions documented. Your team can interrogate every input.
Measurement infrastructure
Fix the data layer that everything else depends on.
Attribution is only as good as the data underneath it. We audit and fix the tracking foundation:
- Tag architecture
- Event schemas
- Conversion configuration
- Server-side tracking
- CRM ingestion
We find the issues that silently corrupt attribution and deliver a clean architecture your team can build on.
Tracking that looks fine until someone looks closely. Duplicate conversions, misfiring pixels, server-side setup dormant for months.
Clean, audited tracking architecture. Deduplicated conversions, validated pixels, working server-side stream, CRM ingestion that actually closes the loop.
MMM, incrementality and signal engineering readiness
Build the foundation that makes advanced measurement possible.
Teams come to us when they want to run Media Mix Modeling or geo-based incrementality tests and need someone to explain what that actually requires, then build it.
That means building the tables, pipelines and data structures that make these techniques possible. We assess and build across:
- Geo-level data quality and coverage
- Regional spend consistency
- Identity stitching
- Conversion feedback loops
- Signal engineering for value-based bidding
We don't just tell you what's missing. We build what you need to get there.
Bidding CPA on a single conversion event. Platforms optimising for who converts cheapest, pulling in low-quality users.
Value-based signals flowing to platforms. Identity stitched across sessions, geo coverage clean, conversion feedback loops in place. MMM and incrementality testing become possible.
Attribution projects designed for your data team to own, maintain and build on top of.
We do not ask you to rip and replace. Everything we deliver integrates with your existing stack.
