GA4 and Google Tag Manager Implementation
Two months is the GA4 default data retention, which means most properties silently cannot compare a year to the one before it. Ivan Sp implements GA4 and Google Tag Manager as part of the initial audit and setup phase of every engagement. Without a working baseline, proving that later SEO or CRO work actually improved anything is impossible, since there is nothing prior to measure against.
Baseline Before Optimization
I establish analytics and heatmap tools during initial setup, before any content or conversion changes begin. A baseline captured after work has already started cannot separate the effect of the work from whatever the site was already doing.
Scoped to What the Client Actually Needs
Implementation gets scoped to the simplest reliable setup that answers a client’s actual questions, rather than the most granular setup technically possible. One funnel build tracked page views with source, medium, country, city, and device attribution, plus a pragmatic form-abandonment calculation using page views minus completions divided by page views.
A more granular field-by-field interaction tracker was available and deliberately skipped, because the client said they did not need it. Overbuilt tracking that never gets analyzed is wasted setup cost. Right-sized tracking that answers real questions gets used.
Third-Party Tool Attribution
Bookings and form submissions handled through a third-party tool, such as a scheduling widget, need enhanced event tagging through GTM to attribute the conversion back to the page that actually produced it. Without this step, a booking shows up in the third-party tool’s own dashboard with no connection to the marketing work that generated it.
Referrer and Attribution Capture
Immediate previous-page referrer data gets captured as part of implementation, showing the specific internal path a visitor took before converting. This detail matters for multi-step journeys, where the page that earned the click and the page that closed the conversion are often different pages entirely.
Implementation Compared to Troubleshooting
This page covers building tracking correctly the first time. Diagnosing why an existing GTM setup is not firing correctly is a distinct problem, covered under Google Tag Manager troubleshooting.
What This Looks Like at Budget Time
Consider a clinic manager presenting channel performance to the partners. Her analytics show the booking tool as the second largest traffic source and organic search producing almost nothing.
She is about to recommend cutting the search budget. She does not want to be wrong about this in front of five people who signed off the spend.
I open the acquisition report and sees the scheduling tool listed as a referral. Every booking passed through it, so every booking was credited to it, and search had been working the whole time.
Where the Standard Implementation Advice Breaks Down
- The usual advice
- Install the tag, confirm real-time shows traffic, and the implementation is done.
- What holds up in practice
- Real-time proves a page view fires, and proves nothing about attribution. The most expensive fault in most setups is a booking credited to the tool that processed it rather than to its real source.
Picture a clinic whose analytics show the scheduling tool as a leading traffic source. Organic search appears to produce almost no bookings.
Every booking passes through that tool, so every booking is credited to it. The budget conversation that follows is based on a number that describes a handover, not a source.
I configure cross-domain measurement and referral exclusions during setup. The bookings were always coming from search, and nothing was recording it.
Attribution faults do not look like faults. They look like a channel underperforming, and budget gets moved away from the thing that was working.
Four Things Commonly Assumed About Implementation
Is your booking tool showing up in analytics as a traffic source?
Ivan Sp meets these four assumptions whenever a setup was installed and never verified.
| Commonly assumed | What actually holds |
|---|---|
| Installing the tag completes the implementation. | The tag records page views. Attribution, cross-domain handovers, and conversion definitions are all separate work. |
| Cross-domain tracking is only for large businesses. | Any site sending visitors to an external booking, payment, or scheduling tool needs it, which is most small businesses. |
| Enhanced measurement should always be fully enabled. | Some components produce noise on some sites. Reviewing and disabling the irrelevant ones keeps the data readable. |
| Data retention can be changed later if needed. | Retention is not retroactive. Data older than the setting is already gone when the change is made. |
| A hard-coded tag is simpler than a container. | A container keeps a version history, allows changes without a developer, and provides the preview mode used to verify a tag. |
A booking credited to your scheduling tool instead of to organic search is how budget gets moved away from the channel that was working. Nothing in the reporting will flag it.
Implementation Reference
- Deployment method
- The GA4 configuration tag, deployed through a Google Tag Manager container.
- Container components
- Tags, triggers, variables, and the data layer that feeds them values.
- Enhanced measurement
- Scroll to 90 percent depth, outbound clicks, site search, video, downloads, and form interactions.
- Session definition
- A session ends after 30 minutes of inactivity by default.
- Data retention default
- Two months, which should usually be changed to fourteen.
- Key events
- The GA4 name for conversions since 2024. Only actions with business value should be marked.
- Verification surfaces
- Tag Manager preview mode, GA4 DebugView, and Google Tag Assistant.
- Attribution capture
- UTM parameters and referrer values, which are what make source reporting meaningful.
- Consent
- Consent Mode, which adjusts tag behaviour based on the visitor’s stated permission.
A Worked Example: A Booking That Lost Its Source
Consider a clinic taking bookings through an external scheduling tool. The comparison below is illustrative, written to show what enhanced tagging closes. No client engagement is described here.
| Step in the journey | Default setup | After enhanced tagging |
|---|---|---|
| Visitor arrives from organic search | Recorded correctly | Recorded correctly |
| Visitor clicks Book Now | Outbound click recorded | Outbound click recorded |
| Visitor lands on the scheduling tool | New session on a new domain | Same session, cross-domain linked |
| Booking completes | Credited to the scheduling tool as a referral | Credited to organic search |
| Which page produced the booking | Unknown | Recorded |
I fix the third row, and the fourth and fifth rows follow from it. The booking was always happening, and the credit was going to the wrong place.
Untagged cross-domain tracking is a relay race where the baton is dropped at every handover. Every runner ran, and no time is recorded.
The final row is what a marketing budget is decided on. Without it, organic search looks like it produces nothing and the scheduling tool looks like a traffic source.
How an Implementation Runs
I verify at each stage rather than at the end, since a broken tag is cheapest to find immediately.
- Collect the business questions first, and write the measurement plan before touching a container.
- Create the Tag Manager container, and install the snippet in the site head and body.
- Deploy the GA4 configuration tag, and confirm page_view fires in DebugView.
- Review enhanced measurement, and switch off any component that produces noise on this site.
- Set data retention to fourteen months, so year-on-year comparison remains possible.
- Configure cross-domain measurement for any external booking, payment, or scheduling tool.
- Add referral exclusions for those domains, so a handover does not start a new session.
- Build custom events only for the questions the measurement plan actually listed.
- Mark as key events only the actions carrying real business value.
- Verify every tag in preview mode and DebugView, then publish the container version with a description.
How Implementation Is Scoped
I scope implementation by the number of external tools that need bridging.
- What arrives
- A measurement plan, a published container, a configured property, and verification screenshots.
- Typical duration
- Two to three weeks for a standard site with one external booking tool.
- What the client supplies
- GA4 and Tag Manager access, site access, and admin access to external tools.
- Included
- Planning, deployment, enhanced measurement tuning, cross-domain setup, verification, and documentation.
- Not included
- Server-side tagging infrastructure, and consent platform licensing or legal advice.
- What changes the estimate
- Ecommerce, several domains, or a booking tool with no tagging support.
- How it is quoted
- After a short scoping call, once the external tools are listed.
Tools and Primary Sources
The references below document each setting named above. Each opens in a new tab.
- Google Tag Manager is the container interface used for deployment and preview testing.
- GA4 enhanced measurement lists which interactions are collected automatically and how to disable each.
- GA4 cross-domain measurement documents the configuration that keeps a session intact across two domains.
- GA4 DebugView shows events arriving in real time, which is how a tag is verified rather than assumed.
- Google Consent Mode documents how tag behaviour adjusts to a visitor’s consent state.
Verify Your Own Tracking
I suggest a verification anyone with GA4 access can run.
- Open GA4, then Admin, then DebugView, and leave it open in one tab.
- In another tab, open your site and complete your most important conversion action.
- Watch DebugView for the event, and confirm it arrives within a few seconds.
- Check the parameters attached to it, and confirm the source and medium are what you expected.
An event that never arrives is a conversion the business has been reporting on faith. A source of direct on a visit that came from search is a cross-domain or referral exclusion problem.
Common Questions
How long does a typical GA4 and GTM implementation take?
A standard setup with a small number of custom events usually completes within a week. A complex, multi-step funnel with third-party tool integration takes longer, scoped during the initial audit.
Does every client need the same tracking setup?
No. Scope follows what the client actually wants to know, confirmed directly with them, rather than a default template applied to every site regardless of what questions matter.
Can tracking be added incrementally after initial launch?
Yes, and it usually should be. Starting with the essential events and adding more as real questions arise avoids building tracking nobody ends up using.
Is Google Tag Manager necessary, or can the tag go straight into the site?
A hard-coded tag works. A container makes every later change possible without a developer, and it keeps a version history. A container also provides the preview mode used to verify a tag.
Discuss This Work
The first call is a free 30-minute consultation with no pitch attached. You receive an honest assessment of where the site stands and what can realistically be built.