Most GA4 Setups Undercount. Mine Reconcile.
Your agency's report says one thing, your CRM says another, and nobody in the room trusts either. Decisions made on that data are guesses wearing charts. I build lead tracking as an engineering discipline, and the acceptance test is that analytics and the CRM agree.
What I Actually Do to Your Tracking
GA4 that survives the real web
Ad blockers, consent banners, and Safari eat client-side tracking. I configure GA4 with server-side tagging where it earns its keep, so the counts you report are close to the leads that actually happened.
GTM without the tag soup
One container, documented, with a naming convention a stranger could follow. Every tag has an owner and a reason. When something double-fires or silently dies, you find out from me, not from a quarterly report that looks wrong.
Reconciled to the CRM
The test is simple: the leads in your analytics match the leads in your CRM, by name and by source. When the two disagree, I treat it as a bug and fix the wiring until they agree.
Dark funnel, handled honestly
Slack shares, podcasts, and word of mouth never show up in last-click reports. I pair the tracking with self-reported attribution so the sources analytics cannot see still get counted.
Reporting you can act on
Not a 40-tab dashboard. A short view of which sources filled the pipeline, what each cost, and what to do next month because of it. If a number cannot change a decision, it does not make the report.
My Take: Attribution Is Trust Architecture
When analytics and the CRM disagree, the loudest opinion in the room wins the budget, and marketing gets blamed for whatever happens next. The fix is not another dashboard. It is wiring the measurement so both systems tell the same story, then being honest about the sources no tool can see. I have written about what should replace last-click, including the dark-funnel sources no tool can see. The promise block further down this page is the short version; this whole page is the long one.
Read the full argument: GA4 vs server-side tracking →This Is Built, Not Promised
This is the part of the work I dogfood hardest. The tracking on the site you are reading is the same package I install for clients, and the reporting pilot runs on real client pipelines today. As those engagements mature, the reconciled before and after numbers will be published right here.
Go Deeper
Beyond Last Click: B2B Attribution Models That Work and the Ones That Lie
The full architecture: what connects to what, and in which order to build it.
Read the post →GA4 vs Server-Side Tracking: Which Numbers Can You Trust?
Why owning your measurement matters more every year third-party data erodes.
Read the post →HubSpot Setup: The Wiring Checklist That Makes It Worth Paying For
A structure for turning tracked data into decisions instead of dashboards.
Read the post →Reported Against Pipeline, Not Activity
Everything I run is tracked and reported against the numbers your leadership team already cares about: qualified pipeline, cost per opportunity, closed-won revenue. You get a live dashboard plus a reporting cadence we set up front, so you see what was done, what it moved, and what it cost. And when a channel can't earn its budget, I say so.
Want tracking your CRM and your CFO both believe?
Running a larger growth engagement? See the Fractional Growth Partner engagement →