I worked on one project. After a few months it is clear: users take long paths and return later, and the 30 minute limit cuts the thread. We lose track of where the person is in the journey.
As a result attribution breaks: some returns count as new sessions, conversions and ROAS drop, and the retargeting budget goes to waste. Clients see numbers that don’t make sense and we scramble to explain why.
Why does this wreck attribution and what about the budget?
• Breaks in the chain mean repeat visits aren’t counted properly; we understate the impact of returning customers.
• As a result conversions and ROAS look lower and the retargeting spend goes to people whose signals don’t align with our analytics.
How to align session windows and stitch sessions via server side tracking?
• Implement server side tracking with a single FPID, so we stitch sessions on the server and aren’t limited by browser timeouts.
• Rely on custom session table in BigQuery.
What steps to take and how to verify impact?
• Check and adjust timeouts in GA4 and ad platforms as needed.
• Run a pilot and analyze results in BigQuery - watch conversions, ROAS and the share of repeat visits.
Want to get all my top Linkedin content? I regularly upload it to one Notion doc.
Go here to download it for FREE.
