PRACTICE 03 · ANALYTICS
Marketing Analytics
Most marketing orgs measure clicks, not pipeline. We build the measurement infrastructure that turns marketing spend into a line item the CFO trusts — attribution, incrementality, dashboards, and the operating cadence that keeps the data honest.
ENGAGEMENT SHAPE
Six-week diagnostic → multi-quarter implementation
Typical first engagement: a six-week analytics diagnostic, a written point of view on the current measurement stack, and a multi-quarter implementation retainer. We work in your data, in your stack, with your team.
WHAT WE DO
The work, in concrete terms.
Attribution model design: MMM, MTA, geo-test, or hybrid — chosen by the question
Dashboard architecture: the four-to-six reports the leadership team actually reads
Data pipeline design: the events, sources, and joins that produce the numbers
Incrementality testing program: quarterly calendar, methodology, writeup
CFO-facing reporting: the slide that gets marketing a re-forecasted budget
Internal training: the marketing team learns to defend the numbers themselves
WHY IT MATTERS
The mechanism, not the artifact.
Attribution is a precondition for marketing to scale.
If you cannot trace spend to pipeline, you cannot defend the budget. The first job of analytics is to make marketing a line item the CFO trusts at the budget meeting, not just a quarterly report.
The model is a function of the question.
MMM is right when you are allocating budget across channels. MTA is right when you are optimizing within a channel. Geo-test is right when you are measuring a category launch. We choose the model that answers the actual question, not the one that is currently fashionable.
Dashboards do not change behavior — the cadence does.
A dashboard nobody reads is a dashboard that does not exist. The work is in the cadence: who reviews the data, on what day, with what decision rights, and what the response is when a number moves.
DELIVERABLES
What you receive.
- 01A 25–40 page analytics diagnostic with a written point of view on the current stack
- 02An attribution model implemented in your data, with named sources, named methods, named owners
- 03A 4–6 report dashboard for the leadership team, with a review cadence
- 04A incrementality testing program: methodology, calendar, last-cycle results
- 05A CFO-facing reporting template, with the slide that survives the budget meeting
- 06An internal training program: 8–12 hours, with the team learning to defend the numbers
OUTCOMES WE TRACK
How success is measured.
- A working attribution model in your stack, with named owners and a documented methodology
- A 4–6 report dashboard the leadership team reads weekly
- A defensible incrementality testing program that runs quarterly
- A documented methodology the CFO can audit
- A marketing team that can defend the numbers themselves
PRACTICE FAQ
On this practice specifically.
Do you require a particular vendor?
No. We work in your stack, with your data, with your team. We have built attribution in Amplitude, Mixpanel, mParticle, Segment, BigQuery, Snowflake, Databricks, and a dozen homegrown stacks. The right tool is the one your team will use.
Do you build the dashboards, or do you just specify them?
We build them, in your BI tool. We do not deliver a spec and walk away. We deliver a working dashboard, with the data pipeline, with named owners for maintenance, and an internal training program so the team can extend it.
NEXT STEP
Bring us the brief.
Six-week diagnostic, NDA from day one, written point of view. We will tell you in the first call whether we are the right firm for the problem.