Attribution simulator
Shows how one missing link moves credit between campaigns
Measuring demand gen
The pipeline is there, but nobody knows which campaign produced it.
Goal
Know which channels and campaigns produce pipeline per euro spent, state the method and its limits, and decide accordingly.
Key metric
Share of opportunities with a tracked source; sourced pipeline per euro spent.
Common mistake
One attribution model presented as the truth, read by contact instead of by account, and optimised for leads and clicks instead of opportunities.
Who owns it
Marketing ops and RevOps
Where AI fits in
A typical case: the step in this phase where AI changes the work most, and who does what.
The monthly campaign review
Generative AI- 1 · CollectToolsn8n and SalesforcePulls the month's campaigns, costs and pipeline
- 2 · CheckToolsn8nCounts records with no campaign or no contact roleAIClaudeFlags data quality issues and how they change the reading
- 3 · ReadAIClaude, with the skillWrites the review: what changed, hypotheses, actions
- 4 · DiscussPeopleMarketing and salesCheck the hypotheses in the meeting
- 5 · DecidePeopleHead of demand genDecides budget, targeting and testsToolsHubSpotBudget and lists updated
BeforeA day matching up spreadsheets; the meeting argues about the numbers.
AfterAn hour of checking; the meeting discusses decisions.
The limitAI suggests hypotheses, not causes: a person checks them. With few cases, the numbers stay unstable.
Indicative times, based on a typical case.
How it works
- Set the rules: what counts as an opportunity, the lookback window, sourced and influenced creditMarketing and sales
- Check data quality before every reportMarketing ops
- Rebuild the history of each opportunity and assign creditMarketing ops
- Read by channel, campaign and cohort: cost per opportunity, pipeline per euroMarketing ops
- Compare the tracked source with the one the customer reportsMarketing ops and sales
- Turn hypotheses into tests, one variable at a timeDemand generation
- Decide in the review: budget, targeting and testsMarketing, SDRs and sales
Resources
Campaign measurement
Sourced and influenced pipeline, cost, and pipeline per euro spent, by campaign and cohort
Campaign performance review
Reads the campaign export and writes the commentary: what changed and what to do
Data quality before the report
Checks to run before the monthly report: leads with no campaign, opportunities with no contact role
How to install the skill: in Claude, upload the zip in Settings, under Skills. In Claude Code, unzip the folder into ~/.claude/skills/. Then just ask: "Here's the campaign export. Please prepare the review."
Nordwind Logistics
Example company · 3 people in the buying group
Opportunity created · €80,000Example company · 3 people in the buying group
The journey in the 90 days before the opportunity
The three links
Attribution model
Attributed pipeline
No source
Source campaign
Credit by campaign · dashed line = value with all links active
- Interactions with no source campaign · threshold 5%
- People not linked to an account · threshold 10%
- Opportunities with no contact role · threshold 10%
- Opportunities with no source campaign · threshold 20%
- Active campaigns with no cost · threshold 0%
- Moves to MQL or SQL with no date · threshold 5%
- Rejected leads with no coded reason · threshold 5%
Building AI into your demand gen?
Share what works for you, or ask about a resource.