Acquisition analytics · GA4
Audited 60K+ GA4 users to separate real, engaged traffic from noise, corrected an inflated reach metric, and pointed acquisition spend at the channel that actually engaged.
The question
Reach and session counts flatter a report. The task was to test whether the audience was as large as it looked, and then to find which acquisition channels brought users who actually engaged rather than users who merely arrived.
The correction
Segmenting by session depth surfaced a block of sessions with near-zero depth, consistent with non-human traffic. Removing it corrected a materially inflated reach number that acquisition decisions were being made against.
Reported audience, after the audit
segmented by session depth · GA4
Isolating the 10% stopped acquisition investment from being sized against a reach figure that was not real.
The signal
Engagement by acquisition channel
GA4 acquisition + landing-page analysis
Organic search delivered roughly 3× the engagement of other channels, which reshaped where acquisition effort should go.
The call
Reallocate acquisition toward organic search, and report reach on the cleaned audience. The cheapest win here was not a new channel, it was refusing to pay for traffic that never engaged.
How I'd measure it
Limits and next steps
Confirm the near-zero-depth block is truly non-human, not a real low-intent segment worth keeping.
Test whether organic's engagement edge holds as volume grows, or thins at scale.
Check the channel comparison against a fair lookback window, not last-click credit.
Confirm engaged organic sessions convert, so engagement is a proxy for value, not vanity.