Apply filters to views

I need the three reporting views for our storefront set up so analysts stop pulling raw data. Create a view that excludes internal traffic from IPs 203.0.113.5 and 203.0.113.18, another that includes only paid-channel sessions, and a third that contains only transactions from our US store. Verify the filters by comparing session counts to the unfiltered master view and tell me which filter order you used.

Before I hand this filtered view to the weekly reporting team, make the user-facing metrics simple: surface the difference in sessions from the master view at the top, label which filters removed traffic, and flag if any filter would drop more than 15% of sessions. Also highlight any page paths that disappear entirely when the view is applied.

I just added the exclude-internal-IP filter and web sessions dropped by 28% for this view; the operations lead Tom will assume we broke tracking. I do not have server logs to validate and I cannot tell whether those IPs are larger ISP ranges or actual office desks. What likely caused the plunge and what should I do next to prove the filter is correct or roll it back?

Every quarter I recreate the internal-excluded, paid-only, and US-transactions views because filters get lost or misordered, and that wastes days and erodes trust. Which habit change will stop this — a checklist, a canonical view state, or a different naming and verification routine — and give the exact steps I should take now to make it stick?