Fractional measurement team and managed infrastructure

Your marketing moves faster than your data team can measure it

You get a team that works out what your marketing needs to measure, builds it, runs it, and reads the results with you. So you know where to spend, and how much. No data team needed.

Book a call

Budget calls are hard to make on numbers you don’t trust

Most teams we talk to already know where their reporting is off. What they’re missing is the time and the people to fix it.

01

Some channels take credit for others

Search and direct get credit for demand that paid social, podcasts or video created. The channels doing the work look weaker than they are.

02

Launches get measured after the fact

A new channel goes live, and weeks later the team tries to work out whether it worked, with no agreed way to tell.

03

Data requests wait in a queue

Your data team also works for product, finance and the rest of the company. Marketing questions wait their turn.

Your first month is a baseline, then we build and analyse

Month one works like an audit. After that, we build what’s missing and improve the setup as your marketing changes.

Month 1

The baseline

What’s in place, what’s running, and what needs building first.

You get: a plan for what to build first
From month 2

Build and analyse

One view of every customer, attribution rules and tests, with a budget read every month and direction every quarter.

You get: numbers you can make budget decisions on

You know how each launch will be measured before it goes live

We work on a set rhythm with your marketing team, so measurement keeps up with what you launch.

01

Before a launch

We agree what it should move, what we’ll look at, and when we’ll first read it.

You get: a measurement plan
02

After a launch

We compare what showed up in revenue with what we expected.

You get: a readout
03

Every month

Where we’d move budget, and where the numbers are least reliable.

You get: a budget read
04

Every quarter

Direction on budget, and which tests to run next.

You get: a plan for the quarter

What a budget read looks like

For each channel: what the evidence says, how sure we are, and what we’d do. Where the numbers are weak, we say so. We walk you through it on a call.

Budget readIllustrative example
ChannelWhat the evidence saysHow sureWhat we’d do
Paid search Clicks and survey answers agree High Keep budget steady
Meta Few clicks, often named in the survey Medium Test before scaling
Podcasts Almost no clicks, named by new customers Low Run a test next quarter

Each round, more of your budget sits on numbers you can use

The budget read shows where the numbers are weakest. That decides what we build next: a new data source, a changed rule, a test, or a tool you don’t have yet, like MMM or an MMP. We help you decide which methods and tools you need, and when.

Then the next round starts

What we build for you

You don’t need a warehouse or any data setup to start. We build the marketing data foundation for you, and add to it as your needs change.

One view of every customer, from first touch to revenue

Ad clicks, website and app visits, CRM deals, revenue and survey answers, joined into one record per customer. You see the whole path to revenue, including the touchpoints a last-click report leaves out.

Customer journeyIllustrative example
  1. Day 1 Clicked a LinkedIn ad Ad platform
  2. Day 1 Read two blog posts Website
  3. Day 9 Came back by searching the brand name Website
  4. Day 9 Started a free trial Product
  5. Day 9 Answered “A podcast” in the signup survey Survey
  6. Day 23 Became a paying customer, $2,400 a year Billing

Last-click report: branded search

This journey: LinkedIn ad, blog, podcast, branded search

Attribution rules your marketing team can read and change

Which traffic counts as which channel, and which touchpoints get credit, written in plain language. When you launch a new campaign or partner, you update the rules yourselves. When someone asks why a channel got credit, the answer is in the rule.

RulesLast changed by your marketing team, 3 days ago
Channel groups
  • Source contains podcast counts as Podcasts
  • Medium is partner counts as Partners
Credit
  • When paid search follows another touchpoint, split the credit between them
  • When the survey names a channel, give it a share of the credit
  • Touchpoints more than 90 days before signup get no credit

Clicks, surveys and tests, read together

To know which channels bring customers you wouldn’t get otherwise, one source isn’t enough. Some channels leave clicks, some show up in surveys, and some only show up in tests. We read all three, with rules for which one leads when. We call this triangulated attribution.

When the data can’t settle a question, we design an incrementality test with you: where to run it, the budget and how long.

Answers to your questions in Claude

Your team asks questions in plain English and gets answers from the same data and definitions as your reports. Answers show how they were calculated, and we fix the ones that come back wrong in our working calls.

Illustrative example

Ad platforms that bid on the customers you want

The customer journey we build for your reporting already holds what ad platforms need: click IDs, conversions and revenue. So we use the same infrastructure to tell Meta, Google and LinkedIn which customers are worth finding, and they bid on more people like them.

We decide what each platform gets, check what was sent against what came back, and test whether your return improved. More in our signal engineering workshop recap.

Same infrastructure, a second use

We’ve built this many times, so you get there faster

What usually takes 2 to 3 hires
  • Measurement strategy
  • Budget allocation
  • Data modelling
  • Analytics engineering

In one team.

We start from a foundation

We adapt what we’ve already built to your data, instead of starting from zero. AI speeds up the models and ingestion, based on our past builds.

We’ve solved these problems before

Less trial and error. Both founders have around 15 years in marketing measurement and data.

We start with what matters

We find the biggest gaps early, so the first months go to what affects your budget most.

Who you work with

Barbara Galiza

Barbara Galiza

Marketing measurement

Growth lead for mobile apps, then around ten years as an independent consultant, with brands like Microsoft, WeTransfer and VEED. Writes the 021 Newsletter.

Timo Dechau

Timo Dechau

Data engineering

Builds the data systems marketing teams rely on, for companies like Amplitude, Shimano, Cricut and Audible.

Barbara and Timo work on every project. We take on at most six clients at a time.

Is it a fit?

A good fit if

  • Your attribution misses a big part of your marketing, because of the channels you run or how your customers buy
  • Your channel mix is changing, or you want to spend more and need to know where
  • You launch new channels and campaigns often
  • You don’t have a data team, or it doesn’t have time for marketing

Probably not a fit if

  • You only run paid search
  • You’re looking for a tool
  • Your data team already covers marketing measurement well
  • There’s no budget decision riding on the answers
  • Your media budget is small. Measurement should cost a small share of the spend it steers

Managed infrastructure, without being locked in

Everything runs in our infrastructure, and we keep it running: monitoring, fixes, and new sources as you need them. It’s monthly, with no minimum term. If you stop, you take everything with you.

For every setup, we share a DPA, a security overview and our GDPR basis.

Questions teams ask before starting

Do we need a data team?

No. If you have one, they keep their platform, and we take the marketing work off their backlog. We can load from their platform and feed results back into it.

How long until we see the first numbers?

About a month for the first standard data sources, two if they’re custom or there are many.

What do you need from us?

A workshop with marketing and one with data, access to the tools in scope, and time for the working calls.

How is this different from an attribution tool?

A tool gives you its model and asks you to fit around it, and someone on your team still has to run it. We design the measurement around your marketing, and run it for you.

Can you tell us exactly how much revenue each channel drove?

Not exactly. We show where each channel appears in the journeys that turn into revenue, and use surveys and tests where clicks don’t reach. That’s usually enough for a much better budget call.

Do you run campaigns or set up tags?

No. We don’t run campaigns, add tags or pixels to your site, or build predictive models like lead scoring. We work out what needs tracking, and your team or agency implements it.

The first call is for us to work out whether Propel is the right fit.

If we think it is, we book a second call. If we don’t, we’ll tell you.

01

First call: is it a fit?

A 30-minute diagnostic. We ask about your channels, budget, how you measure today and what’s changed.

02

If we think it’s a fit, we do the homework

The team goes through what you shared, works out where your measurement is likely falling short, and puts together our recommendations.

03

Second call: our insights and our take

A few days later, we walk you through what we found, what we’d do next and why, the scope and the price. It should be useful even if you don’t work with us.

Propel