GA4 Setup Services and Google Analytics 4 Audits

GA4 records every interaction as an event with parameters, rather than organising data around sessions, pageviews and category-action-label events. It combines website and app data in one property, uses key events instead of goals, relies more on machine learning and modelling, offers flexible exploration reports and provides free BigQuery export. Reports and metrics are defined differently, so numbers are not directly comparable with the old platform.

A measurement plan lists the business questions you want to answer, the metrics that answer them, and the events, parameters and dimensions needed to calculate those metrics. It prevents both under-tracking, where important actions are missing, and over-tracking, where hundreds of unused events clutter reports. It also documents naming conventions so future tracking stays consistent.

Beyond the events GA4 collects automatically, most websites should track actions that show intent and value.

Key events, previously called conversions in GA4, are the events you mark as most important to the business, such as purchases or qualified lead submissions. They appear in reports and can be imported into Google Ads for bidding. Marking too many events as key events dilutes their meaning, so we select a small set that reflects real business value, with values attached where possible.

GA4 has recommended ecommerce events with a standard structure for items, prices, quantities, currencies, coupons and transaction IDs. Correct implementation, usually through a data layer provided by your store platform or developers, enables product, category, funnel and revenue reports. Transaction IDs prevent duplicate purchases when customers refresh confirmation pages. Shopify, WooCommerce and Magento each have integration options with different strengths and limitations.

When customers leave your site to pay through a gateway and return to a confirmation page, GA4 may treat the gateway as the source of the visit, crediting sales to the payment provider rather than the ad, search or email that brought the customer. Adding payment domains to the list of unwanted referrals, and checking cross-domain configuration, keeps attribution with the real channels.

Consent mode lets Google tags adjust their behaviour based on visitors’ cookie choices. When consent is denied, tags send limited signals without storing cookies, and Google can use modelling to estimate some missing data. For businesses serving visitors in regions with strict consent rules, and for Google Ads features that rely on consent signals, consent mode is increasingly important. It must be connected correctly to your cookie banner. In India, the Digital Personal Data Protection Act also sets expectations for consent and transparency.

Visits from your own team, developers and agencies can distort data, particularly on smaller websites. GA4 lets you define internal traffic by IP address and filter it from reports, and developer traffic can be excluded separately. Because many teams work remotely or use changing IP addresses, we sometimes use additional methods such as tagging internal users through Tag Manager.

Linking GA4 with Google Ads allows key events to be imported as conversions, audiences to be shared for remarketing and ad performance to be analysed alongside site behaviour. Linking Search Console adds organic search queries and landing page data. Linking Merchant Center and BigQuery extends ecommerce and analysis capabilities. We configure these links and check that conversions are not double-counted across platforms.

BigQuery export provides raw, event-level data that is not subject to the sampling, thresholds and retention limits of standard GA4 reports. It allows advanced analysis, joining with CRM or sales data and building custom attribution or customer lifetime value models. The export is free to set up for standard properties, with BigQuery usage costs depending on data volume and queries. For growing businesses, starting the export early preserves history.

UTM parameters tag links in emails, social posts, WhatsApp messages, QR codes and some ads, so GA4 can attribute visits to the right source, medium and campaign. Inconsistent tagging, such as mixing “facebook”, “Facebook” and “fb”, fragments reports. We create a UTM naming convention and a simple builder sheet so everyone tags links the same way.

A GA4 audit reviews property settings, data streams, enhanced measurement, event naming and duplication, key events and values, ecommerce data accuracy, referral exclusions, cross-domain setup, internal traffic filters, consent mode, data retention, integrations, custom definitions, UTM consistency and report usability. Findings are prioritised by impact on decisions, with clear instructions for fixes.

These issues appear in many GA4 properties.

GA4 uses attribution models to assign credit for key events to marketing touchpoints. Data-driven attribution distributes credit based on how different touchpoints contribute to conversions in your data, while last-click gives all credit to the final touchpoint. Reports and Google Ads conversions may use different models and windows, so we explain how to read them. No attribution model is perfect; comparing channels over time and running tests gives a more reliable picture.

GA4 rarely matches ad platforms, CRMs or store backends exactly. Differences come from consent choices, ad blockers, attribution models and windows, time zones, definitions of users and sessions, and data processing delays. The goal is not identical numbers, but understanding the size and cause of differences. Large gaps usually indicate tracking problems worth fixing. Server-side tracking can improve data completeness.

Standard GA4 reports can feel unfamiliar. We customise the report navigation around your key questions, build explorations for funnels, paths and segments, create audiences for analysis and remarketing, and train your team on finding answers. For management reporting, Looker Studio dashboards combine GA4 with ads, CRM and sales data in a simpler view.

Yes. GA4 properties can include app data streams through Firebase, alongside website data, giving a combined view of users across web and app. App tracking requires the Firebase SDK in your Android and iOS apps and consistent event naming with the website where possible. Our app development team can implement app analytics.

In India, many enquiries happen by phone or WhatsApp rather than forms. Clicks on phone numbers and WhatsApp buttons can be tracked as events and marked as key events, with parameters showing which page and button were used. For a fuller picture, call tracking numbers and WhatsApp Business Platform integrations can pass completed calls and conversations back to analytics or the CRM.

GA4’s default retention for detailed event data used in explorations is short. Increasing it to the maximum available for your property type preserves more history for analysis. Standard aggregated reports are not affected by this setting, but explorations and custom analyses are. BigQuery export keeps raw data for as long as you choose.

Costs depend on the number of websites, apps and domains, the complexity of events and funnels, ecommerce requirements, consent setup, integrations, BigQuery export, data layer development, validation effort and training. Audits are usually a fixed fee based on property complexity.