GA4 inaccuracies almost always come from three places: incorrect event configuration (duplicate events, missing parameters, misfired GTM triggers), consent mode issues where users declining cookies are not modelled correctly, or discrepancies between GA4 and ad platform data due to different attribution windows. A systematic audit typically resolves 80–90% of data quality issues quickly.
Server-side tagging moves data collection from the user’s browser to a server you control. Ad blockers and browser restrictions no longer interfere, page speed improves with fewer browser scripts, and first-party data collection becomes more reliable. If you run paid media at meaningful scale and care about accurate attribution, server-side tagging is worth implementing.
Universal Analytics used a session and pageview model with predefined hit types. GA4 uses an event-based model where everything is an event with customisable parameters, built for cross-device tracking and native Google Ads integration. Universal Analytics is no longer processing data — if you still rely on it for decisions, migration is urgent.
Each platform has its own conversion tracking (Google Ads uses GA4 import or the global site tag, Meta uses Pixel or CAPI, LinkedIn uses the Insight Tag). Without deduplication, the same conversion appears in every platform’s report, inflating attributed revenue by 2–4x. We implement a single measurement plan with GA4 as the source of truth, imported correctly into each platform.
Consent mode is a Google framework that limits data collection when users decline cookies, using modelling to fill gaps. Consent mode v2 became mandatory for EEA advertisers in March 2024. Without it, your Google Ads campaigns lose access to conversion modelling and your GA4 data significantly under-reports for users who decline — a large proportion of European traffic.
Platform-reported attribution is always inflated — every platform takes credit for every conversion it touched. The most reliable approach combines GA4 data-driven attribution as your baseline, consistent UTM tagging across all campaigns, and channel-level incrementality testing where feasible. We build Looker Studio dashboards that show contribution across the funnel, not just last-click wins.
Yes, in most cases. Our process starts with a full audit before recommending any rebuild. Many problems — duplicate events, broken triggers, incorrect conversion settings — can be resolved without touching your site code. A full rebuild is only recommended when the existing setup is so tangled that fixing individual issues creates more instability than starting clean.