B2B attribution is hard because the data lives in three different places — your ad platforms, your CRM, and your web analytics. Each source gives you a different answer about which channel is working. This stack pulls those sources together, aligns them on a common definition of credit, and runs the combined dataset through a Claude attribution audit prompt. The output is a structured attribution report that shows which channels and touchpoints are actually driving pipeline — not just impressions or clicks.
The Stack
The Prompt
This stack is built around the Attribution Audit Diagnostic Prompt. Here's the abbreviated version — the full prompt with all variables and usage notes is on its own page.
You are a B2B marketing attribution analyst auditing a multi-channel demand gen program. Review the touchpoint and pipeline data provided below. Identify the primary attribution gaps — channels with strong upper-funnel data but weak CRM tracking. Classify each channel by attribution confidence: High, Medium, or Low. For channels with Low confidence, identify the specific tracking failure causing the gap. Output a structured attribution report with a prioritized fix list.[ ... continued — see full prompt ]
The Workflow
-
Export CRM pipeline data
Pull closed-won and active opportunity data from Salesforce — including campaign influence, lead source, and any UTM data captured at conversion.
-
Export GA4 session data
Pull channel-level session data from GA4 for the same period. Include source/medium breakdown and any goal completions tied to conversion events.
-
Align the datasets
Map Salesforce lead sources to GA4 source/medium values. Identify discrepancies — deals with no GA4 session match, or sessions with no CRM record.
-
Run the attribution audit
Paste the aligned data into the Attribution Audit Diagnostic prompt. Claude will classify channels by confidence and surface the specific tracking failures causing gaps.
-
Document and act
Export the attribution report to Sheets. Log the fix list in Notion with owners and due dates. Present the model and gaps to leadership before the next planning cycle.
What This Replaces
- Manual spreadsheet attribution models that take a full day to build
- Ad platform reporting that overcounts conversions by ignoring the full buying journey
- CRM attribution that only captures last touch and misses 60% of the channel story
Related Stacks
New stacks drop weekly.
Each one includes the tools, the Claude prompt, and the workflow logic. Free — built for in-house B2B demand gen managers.