Skip to content

Business · 9 min read

How to Build a GA4 Measurement Plan (Events, Conversions, Consent Mode and Server-Side Tagging)

A step-by-step guide to planning GA4 events, key events, consent mode and server-side tagging so your analytics data is consistent and trustworthy.

Performance analytics graphs on a laptop screen
On this page 13 sections
  1. Why GA4 needs a measurement plan
  2. Step 1: Start with business objectives, not tags
  3. Step 2: Understand GA4’s four types of events
  4. Step 3: Map actions to recommended events first
  5. Step 4: Define a naming convention and respect GA4 limits
  6. Step 5: Choose key events carefully
  7. Step 6: Build the plan document
  8. Step 7: Plan for consent mode
  9. Step 8: Decide whether server-side tagging is worth it
  10. Step 9: Governance, privacy and data quality
  11. Common measurement plan mistakes
  12. The bottom line
  13. FAQs

A GA4 measurement plan is a written document that links your business objectives to the specific events, parameters and key events you will collect, plus the rules for consent, data quality and ownership. To build one, define objectives and KPIs, map them to GA4’s event model, standardize naming, mark only true outcomes as key events, then decide how consent mode and server-side tagging fit.

Without a plan, GA4 tends to fill up with inconsistently named events, duplicate conversions and reports nobody trusts. With one, marketing, product and engineering share a single definition of what success looks like and how it is measured. This guide walks through each step, with examples you can adapt.

Why GA4 needs a measurement plan

GA4 is built around events. Every interaction, from a page view to a purchase, is recorded as an event with optional parameters. That model is flexible, but flexibility without rules creates problems quickly:

  • Different teams create events for the same action with different names, so totals never match.
  • Too many events are marked as key events, so reports stop signaling what matters.
  • Parameters are collected but never registered as custom dimensions, so they cannot be used in standard reports.
  • Personal data leaks into URLs or event parameters, creating policy and privacy risk.
  • Consent choices are handled inconsistently across tags.

A measurement plan prevents these issues by deciding the rules before implementation, and it becomes the reference document for anyone who touches tracking later.

Step 1: Start with business objectives, not tags

Begin with what the business is trying to achieve, then work down to what you need to measure. A simple hierarchy works well:

  1. Business objectives: for example, grow online revenue, generate qualified B2B leads, or increase product sign-ups.
  2. Key performance indicators: the handful of numbers that show progress, such as revenue, conversion rate, qualified leads or activation rate.
  3. User actions: the specific behaviors that drive each KPI, such as adding to cart, starting checkout, submitting a demo request or completing onboarding.
  4. Events and parameters: how each action will be recorded in GA4, with the context needed to analyze it.

Involve the people who will use the data. Sales leaders care about lead quality, finance cares about revenue matching the order system, and product teams care about feature usage. Agreeing on definitions up front, such as what counts as a “qualified” lead, saves painful debates later.

Step 2: Understand GA4’s four types of events

Google groups GA4 events into four categories. Knowing which bucket an action falls into tells you how much implementation work it needs.

Event typeWhat it isImplementation effort
Automatically collectedCollected by default when GA4 is installedNone
Enhanced measurementCollected when enhanced measurement is enabled, such as page views, scrolls, outbound clicks, site search, video engagement, file downloads and form interactionsToggle in settings; verify accuracy
RecommendedEvents you implement using Google’s predefined names and parametersDeveloper or Tag Manager work
CustomEvents you define yourselfDeveloper or Tag Manager work, plus documentation

Two practical rules follow. First, check what enhanced measurement already captures before building anything; for example, its scroll event fires when a user reaches 90% page depth, and its form events may not fit every form setup, so validate them against your site. Second, Google advises that you only create custom events when no other events work for your use case. Recommended events unlock existing and future reporting features.

Google publishes a reference list of recommended events. Using these names consistently is the fastest way to get reliable reports.

Ecommerce

A standard funnel uses view_item, add_to_cart, begin_checkout, add_shipping_info, add_payment_info and purchase, with refund for returns. For purchase, the reference marks transaction_id and items as required, and currency and value are needed for revenue to be computed accurately. The transaction ID matters more than it looks: it is how you reconcile GA4 with your order system and avoid double counting.

Lead generation

Recommended lead events include generate_lead, qualify_lead and close_convert_lead. The first is usually fired on the website; the later stages often happen in your CRM and can be sent to GA4 later, which lets you analyze marketing by lead quality rather than raw form fills.

Accounts and SaaS

sign_up and login are recommended events for account creation and access. Product-specific milestones, such as completing onboarding or inviting a teammate, typically become custom events.

Step 4: Define a naming convention and respect GA4 limits

For anything custom, set rules and write them down. A good convention is lowercase with underscores, follows an object-action pattern such as quote_requested or calculator_completed, and avoids vague names like click or button.

Design within GA4’s documented collection limits. For standard properties these include:

  • Event names up to 40 characters. Google notes that a key event whose name exceeds 40 characters will not be reported as a key event.
  • Up to 25 parameters per event, with parameter names up to 40 characters and most parameter values up to 100 characters.
  • Up to 25 user properties per property.

Custom dimensions and metrics also have quotas: for standard properties, 50 event-scoped custom dimensions, 25 user-scoped, 10 item-scoped and 50 custom metrics. Parameters are only available in standard reports once registered, so decide which ones deserve a slot. Reusing general-purpose parameters, for example a single form_name parameter across all forms, stretches those quotas much further than one parameter per form.

Step 5: Choose key events carefully

Google describes a key event as “an event that measures an action that’s particularly important to the success of your business.” Key events were formerly called conversions in GA4; “conversion” now refers to actions used to measure and optimize ad campaigns, and conversions can be created from key events for use with Google Ads.

Keep key events to genuine outcomes, such as purchases, qualified lead submissions, demo bookings or trial sign-ups. Micro-interactions like scrolling or video plays are useful for analysis, but marking them as key events inflates conversion rates and can mislead bidding if imported into ad platforms. Document for each key event:

  • The exact trigger and any conditions, such as “form submitted successfully, not just clicked.”
  • Whether it carries a value, and how that value is set.
  • Whether it is imported into Google Ads or other platforms.
  • Who owns it and how it will be tested.

Step 6: Build the plan document

A measurement plan does not need to be complex. A shared spreadsheet with one row per event works for most organizations. Here is an illustrative excerpt:

ObjectiveKPIEventKey parametersKey event?Trigger
Grow online revenueRevenue, conversion ratepurchasetransaction_id, value, currency, itemsYesOrder confirmation, once per order
Grow online revenueCart abandonmentbegin_checkoutvalue, currency, itemsNoCheckout page first load
Generate qualified leadsLeads by channelgenerate_leadform_name, lead_typeYesSuccessful form submission
Generate qualified leadsQualified lead ratequalify_leadlead_typeYesCRM status change, sent later
Improve self-serviceContent engagementfile_downloadfile_name, file_extensionNoEnhanced measurement

Add columns for implementation method (hard-coded, Google Tag Manager, server-side), consent category, owner, status and last tested date. The document is only useful if it stays current, so assign someone to maintain it.

Consent handling should be designed in, not bolted on. Google’s consent mode lets your consent banner or consent management platform tell Google tags what the user has agreed to. The main consent types are ad_storage, analytics_storage, ad_user_data and ad_personalization, alongside functionality, personalization and security storage types.

There are two implementation approaches:

  • Basic consent mode: Google tags do not load until the user consents. According to Google, when the user does not consent, no data is transferred to Google, not even the consent status.
  • Advanced consent mode: tags load with default consent states, typically denied, and when consent is denied they send consent state and cookieless measurements. This supports more detailed modeling of the behavior you cannot observe directly.

Which approach is right depends on the laws that apply to you, your risk tolerance and your advertising needs. In the US, state privacy laws vary. California’s CCPA, for example, gives consumers the right to opt out of the sale or sharing of their personal information, and the California Attorney General’s office describes Global Privacy Control as one acceptable opt-out method. If you have visitors from Europe or other regions, different rules apply. Confirm your obligations with privacy counsel; your measurement plan should then record which consent category each tag falls under.

Step 8: Decide whether server-side tagging is worth it

With standard client-side tagging, tags run in the visitor’s browser and send data directly to each vendor. Server-side tagging in Google Tag Manager adds a tagging server you control: the browser sends data to your server container, which processes it and forwards it to tools like GA4.

Google highlights several benefits:

  • Performance: fewer measurement tags on your site means less code running in the browser.
  • Data control: you decide how data is shaped and where it goes, and can strip fields before they reach vendors.
  • First-party context: with a custom subdomain, the server can set cookies that page scripts cannot read.

Server-side tagging works with consent mode. Google’s documentation explains that the Google tag passes consent parameters to the server container, so you only need to set up consent mode in the web container. It does not remove your obligation to respect user choices.

The trade-offs are real. You need hosting (Google recommends Cloud Run, though other environments are supported), monitoring and someone who understands the setup. Hosting costs scale with traffic, so estimate them with your provider before committing. For a small site with a few tags, well-governed client-side tagging is often enough. Server-side tagging makes more sense when you run many marketing tags, have significant ad spend depending on accurate conversion data, or need tighter control over what leaves your site.

Step 9: Governance, privacy and data quality

The last part of the plan covers how data stays trustworthy over time.

  • Keep personal data out. Google’s policies prohibit sending personally identifiable information such as email addresses and phone numbers to Analytics. Common leaks come from URLs, page titles, search boxes and form fields, so check query strings and parameters during QA.
  • Set data retention deliberately. Standard properties can keep user-level and event-level data for 2 or 14 months. Retention affects explorations and funnel reports, not standard aggregated reports. Check your setting so year-over-year exploration is possible if you need it.
  • Consider BigQuery export. GA4 can export raw event data to BigQuery; standard properties’ daily export is limited to 1 million events per day, and storage and processing incur BigQuery charges. Raw data is valuable for joining with CRM or order data and for keeping history beyond retention limits.
  • Test before and after every release. Use Tag Manager preview and GA4 DebugView, and compare GA4 purchase totals against your order system regularly.
  • Control access and changes. Limit who can publish Tag Manager containers and edit key events, and keep a change log.

Common measurement plan mistakes

  • Tracking everything. More events do not mean more insight. Each event should answer a question someone will actually ask.
  • Skipping the transaction ID. Without it, duplicate purchases are hard to detect and reconciliation is guesswork.
  • Treating form clicks as leads. Fire lead events on successful submission, not button clicks.
  • Forgetting to register parameters. Unregistered parameters will not appear in standard reports.
  • Setting consent once and never checking it. Banner changes and new tags can quietly break consent behavior.
  • No owner. Tracking decays with every site release unless someone is accountable.

Measurement is also the foundation for optimization. If you plan to test pages and offers, our guide to a practical conversion rate optimization process shows how this data feeds experiments.

The bottom line

A good GA4 measurement plan starts with business objectives, uses Google’s recommended events wherever possible, keeps key events to real outcomes and treats consent, privacy and data quality as part of the design. Server-side tagging can add performance and control, but it is an investment that should follow a clear need. Written down and maintained, the plan turns GA4 from a stream of events into data your team can make decisions with.

If you want help designing or auditing your analytics, our digital growth team can build a measurement plan around your goals. Contact us to get started.

Sources

FAQ

Questions, answered.

What is a GA4 measurement plan?

A GA4 measurement plan is a document that connects business objectives and KPIs to the specific events, parameters and key events you collect in Google Analytics 4. It also records naming conventions, triggers, consent categories, owners and testing status, so marketing, product and engineering teams share one definition of success and tracking stays consistent over time.

What is the difference between key events and conversions in GA4?

Key events are events you mark in GA4 as especially important to your business, such as purchases or qualified leads, and they appear in Analytics reports. Conversions now refer to actions used to measure and optimize advertising campaigns, and they can be created from key events for use in Google Ads. Key events were previously called conversions in GA4.

How many events can you track in GA4?

Google documents no limit on distinctly named events for web data streams, although app streams have a limit. Each event can carry up to 25 parameters, event names can be up to 40 characters, and standard properties have quotas for custom dimensions and metrics. Rather than tracking everything, collect only events that answer real business questions.

Do I need consent mode for GA4 in the US?

It depends on where your visitors are, which state laws apply to your business and how you use advertising data. Consent mode passes users' choices to Google tags, and some US laws, such as California's CCPA, give consumers rights to opt out of the sale or sharing of personal information. Confirm your obligations with privacy counsel before choosing an approach.

What is the difference between basic and advanced consent mode?

In basic consent mode, Google tags do not load until the user consents, so no data, not even consent status, is sent when consent is refused. In advanced consent mode, tags load with default consent states, usually denied, and send cookieless pings when consent is denied, which supports more detailed modeling of unobserved behavior.

Is server-side tagging worth it for a small business?

Often not at first. Server-side tagging adds a tagging server you must host, monitor and maintain, with costs that scale with traffic. It becomes more worthwhile when you run many marketing tags, rely on accurate conversion data for significant ad spend, or need tighter control over what data leaves your site. Well-governed client-side tagging suits many smaller sites.

Can I send email addresses or names to GA4?

No. Google's policies prohibit sending personally identifiable information such as email addresses, phone numbers or names to Google Analytics. Common accidental leaks come from URLs, page titles, site search terms and form fields, so check query strings and event parameters during testing and remove or redact any personal data before it is collected.

How long does GA4 keep data?

For standard GA4 properties, you can set user-level and event-level data retention to 2 months or 14 months. The setting affects explorations and funnel reports, not standard aggregated reports. If you need longer history or raw event-level analysis, consider exporting GA4 data to BigQuery, keeping in mind that BigQuery storage and processing carry their own charges.

Have something ambitious in mind?

Tell us where you are and where you want to be. We'll come back with honest thinking on how to get there.