Google Analytics

Google Analytics is the most widely used web analytics tool. The standard version is free and shows how many visitors a website has, where they come from and what they do there. Since July 2023, Google Analytics 4 (GA4) has been the only version; the former Universal Analytics was switched off. GA4 collects data via a script (gtag.js or Google Tag Manager) embedded in the website and works with first-party cookies. Every interaction – from page view to purchase – is measured as an event.

Examples of data that can be analysed:

• Demographic information about the audience, such as country, region and language

• Browser and operating system

• Search terms from Google Search – via the link with Google Search Console

• Users and sessions

• Engagement time and engagement rate

• Bounce rate (in GA4: sessions without engagement)

• Device used

• Traffic sources (organic search, ads, referrals, social media, email, AI assistants such as ChatGPT)

• Pages and products with the most views and purchases

• Conversions and e-commerce data such as revenue, average order value and abandoned checkouts

Real-time reports let you follow what is happening on the website right now. The data is very valuable for optimising websites, shops and campaigns. Linking with Google Ads is particularly useful: conversions from GA4 can serve as the basis for automated bidding strategies, and audiences from GA4 can be used for remarketing. Connections with Google Search Console, BigQuery or SEO tools are also possible. GA4 is set up via a tag directly in the source code, via Google Tag Manager or via plugins for shop systems such as Shopware and Shopify.

Custom reports and explorations let you put together exactly the metrics that matter for your business. GA4 is often combined with Looker Studio for clear dashboards.

Accounts and properties in Google Analytics

A Google Analytics account can contain several properties, for example one per website or brand. Within a property, data is collected via data streams – for websites as well as iOS and Android apps. This allows web and app to be analysed in a single view.

Optimising content and channels with Google Analytics

The reports show which pages perform particularly well – measured by engagement and conversions – and which still need optimising. You can also analyse which channels bring visitors who actually buy and which only generate traffic. Specialised tools are used for A/B tests; GA4 supplies them with the data for evaluation.

Keeping an eye on mobile usage

Whether smartphone, tablet or desktop – GA4 shows how users behave and convert depending on the device. In e-commerce in particular, where most visits are mobile, this is crucial for improving shops and campaigns in a targeted way.

E-commerce tracking in Google Analytics 4

For online shops, e-commerce tracking is particularly valuable. It shows the entire purchase process – from product view to cart to purchase – and makes visible where customers drop off. Revenue, order value and the performance of individual products and categories can be analysed, as can the contribution of the individual marketing channels. This requires a clean implementation of the e-commerce events, for example via Google Tag Manager.

Data protection, consent and data gaps

In the EU, Google Analytics may only be used with users' consent; in Switzerland, depending on the setup, transparent information with an option to object is often sufficient. Technically, this is implemented via a consent management tool and Google Consent Mode. If users decline or browsers and ad blockers block tracking, data is missing. GA4 can model part of it, but this doesn't replace real measurement. To reduce data gaps, many shops rely on server-side tracking. Compared with Universal Analytics, GA4 also brings some new features:

• Machine learning for predictions, for example on purchase probability

• Event-based data model instead of sessions and page views

• Flexible explorations for deeper analyses

• Automatically collected events such as scrolls or downloads (enhanced measurement)

• Configurable data retention

• No storage of IP addresses

• Web and app in one property

• Free export to BigQuery

• Data modelling where consent is missing (Consent Mode)