Marketing ops and attribution

If the number has no source, it is not a number.

Reporting that cannot be traced back to a live source is decoration. The job is event tracking, lead scoring, and the spend-to-pipeline join, so a weekly scorecard says something true or says nothing at all.

Book fifteen minutes How we do it ↓

You will recognise this if

How we do it

5 steps, in this order.

  1. 1Agree the eventsA written tracking plan. Names, properties, when they fire. Most broken analytics is a naming problem, not a tooling one.
  2. 2Fix the joinSpend on one side, pipeline on the other, and an identifier that survives the trip between them.
  3. 3Score leads honestlyFit and behaviour kept separate, because a blended score hides which one is doing the work.
  4. 4Automate the scorecardWeekly, from live sources. Any number without a source reads "no data", never an estimate.
  5. 5Delete the restMost dashboards track things nobody has acted on in a year. Removing them makes the remainder trustworthy.

What you end up with

  • A written tracking plan and the implementation
  • The spend-to-pipeline join, working
  • A lead scoring model your sales team can argue with
  • A weekly scorecard that builds itself
  • Fewer dashboards

Questions this answers

What people ask before they call.

How do I connect ad spend to pipeline without an attribution tool?

Keep an identifier that survives from click to deal: UTM values and click IDs stored on the contact, carried to the deal, and joined to the spend exported from each ad platform. A spreadsheet or a small query can do the join. A tool helps later, but the identifier has to exist first.

What should a weekly marketing scorecard include?

A few numbers, each traceable to a live source: spend, qualified leads, pipeline created and where it came from, and what changed since last week. Any number without a source should read no data, never an estimate. The short list of decisions needed from leadership belongs next to it.

Should lead scoring blend fit and intent into one number?

No. Fit says whether the company could buy, and intent says whether it is acting like it will. Blended into one score, a strong fit with no activity looks the same as a poor fit that visits often. Keep two scores so sales can see which one is doing the work.

Why do our marketing dashboards disagree?

Usually because events are named or fired differently across tools, not because one tool is broken. A written tracking plan, with each event's name, properties and trigger agreed once, fixes most of it. Then delete the dashboards nobody has acted on, so the remaining ones can be trusted.

Updated

Copy this into your agent

Find where your data actually breaks

It will trace one customer end to end and show you the exact step where the trail goes cold.

I want to find where my marketing data actually breaks, by tracing one real customer end to end. Work through this with me one step at a time.

1. Ask me to pick one recent closed customer and walk me through every step from their first touch to the closed deal, as far as I actually know it.

2. At each step, ask me what system recorded it and what identifier connected it to the step before. Stop as soon as I cannot answer.

3. Tell me plainly where the trail goes cold, and what it would take to close that specific gap.

4. Ask me which numbers my leadership team looks at weekly, and for each one, whether I could trace it back to a live source right now.

5. Tell me which of those numbers I should stop reporting until they can be traced, and how to say that to leadership without it sounding like I am hiding something.

6. Give me the minimum tracking plan that would let me answer "what did this customer cost to acquire" without a spreadsheet exercise.

Where you are guessing about my stack, say so instead of guessing.
Works in Claude, or whatever you already use.

The other five

PositioningNobody can repeat what you do. LaunchLaunch day is covered. Week two isn't. SEO & contentContent is a backlog nobody owns. Paid acquisitionYou're buying clicks, not customers. Lead generationThere's no repeatable way new leads arrive.

Tell us what you’re trying to move.

Fifteen minutes, no deck, no prep. If it isn’t a fit we’ll say so on the call and point you somewhere better.

Book fifteen minutes →