What is analytics setup?
Analytics setup means instrumenting your application to capture how users behave — which pages they visit, which features they use, where they drop off, and whether they complete the actions that matter to your business. Without this, product decisions are made on instinct rather than evidence.
There are two distinct categories of analytics. Marketing analytics (Google Analytics 4) tracks traffic sources, sessions, and broad conversion goals — it's primarily useful for understanding where users come from and whether your acquisition is working. Product analytics (Mixpanel, Amplitude, PostHog) tracks fine-grained user behaviour — what individual users do, in what sequence, and how that predicts retention. Both serve different purposes and are often used together.
Segment sits between your app and analytics providers as a data routing layer — send events once to Segment, and it forwards them to GA4, Mixpanel, your data warehouse, and any other destination, without changing application code.
When does your app need it?
- You want to know which features are actually used vs which are ignored
- You're running a funnel (onboarding, checkout, activation) and need to know where users are dropping off
- Investors or stakeholders are asking for retention and engagement metrics
- You're making product decisions and want them grounded in data rather than assumption
- You're spending on paid acquisition and need to track whether it's converting
- You're submitting to app stores and need Firebase Analytics integrated for mobile event tracking
How much does it cost?
Analytics setup typically adds 3–5 hours of developer time — roughly from $1,000 AUD.
This covers: SDK installation and configuration, user identification and property tracking, a core set of event definitions (signup, login, key feature interactions, conversion events), integration with consent management for cookie compliance, and verification that events are firing correctly via the debugging tools.
Complexity increases with the number of custom events, if Segment is used as a routing layer, if mobile analytics (Firebase) needs to be integrated alongside web, or if a data warehouse destination (BigQuery, Snowflake) is included.
How it's typically built
Analytics is implemented by adding an SDK (Mixpanel JS, PostHog, or the GA4 gtag) to the application, then calling identify() when a user logs in (attaching their user ID and properties to subsequent events) and track() when meaningful actions occur.
Event naming and property schemas should be defined before instrumentation begins — retrofitting a consistent schema onto an ad hoc event collection is painful. A simple event taxonomy document (event name, trigger, properties) saves significant cleanup later.
For Australian users, GA4 requires cookie consent before the tracking script loads — not just a banner, but actual script gating based on consent state. PostHog can be self-hosted for privacy-sensitive deployments, keeping data in Australia.
For React Native apps, Firebase Analytics is the standard choice for mobile event tracking, with data feeding into Google Analytics or BigQuery.
Questions to ask your developer
- Product analytics or marketing analytics — or both? They answer different questions. GA4 for acquisition and traffic; Mixpanel/PostHog for user behaviour and retention.
- Is Segment worth it for your stack? Segment adds cost and complexity, but pays off if you're routing events to multiple destinations and want to avoid re-instrumentation when you change providers.
- Have you defined your event taxonomy? A consistent naming convention and property schema before instrumentation starts saves significant future cleanup.
- Is your cookie consent properly gating analytics scripts? Displaying a banner while loading tracking scripts regardless of consent is non-compliant.
- Do you need user-level tracking or aggregate-only? User-level tracking (Mixpanel) is more powerful but carries greater privacy obligations.
See also: Feature flags and A/B testing · Technical SEO setup · App cost calculator