How TagNexus transformed digital measurement from isolated tracking requests into a scalable analytics ecosystem - supporting marketing, lead generation, and downstream data reporting at enterprise scale.
Sedgwick
Insurance & Claims
Ongoing
GA4 · GTM · dataLayer
BigQuery + Snowflake
Executive Summary
Every enterprise website collects data. Very few collect data that business teams can consistently trust.
As Sedgwick's digital presence evolved across multiple markets and user journeys, analytics requirements expanded rapidly - spanning marketing campaigns, HubSpot form submissions, lead generation funnels, and new website experiences. Each request arrived in isolation. The result was a fragile tracking setup that was difficult to maintain, impossible to scale, and increasingly unreliable as a source of business intelligence.
Rather than continuing to add tags reactively, TagNexus designed a scalable measurement framework built on structured dataLayer architecture, Google Tag Manager, and Google Analytics 4 - with data flowing downstream into BigQuery and Snowflake for enterprise-level reporting.
The outcome was an analytics ecosystem built for long-term scalability, not short-term fixes - one that marketing, technology, and business intelligence teams could all rely on.
The Challenge
The objective was never simply to deploy GA4 and add tracking tags. The challenge was to create an enterprise measurement strategy capable of supporting continuous website enhancements, multiple marketing programmes, and cross-system data flows - without compromising reporting quality or requiring constant rework.
Tracking was being added request by request with no overarching architecture. Each new campaign or page required a new GTM tag, creating a container that was increasingly difficult to maintain and impossible to audit reliably.
Hidden UTM parameters, form submissions, and marketing attribution needed end-to-end validation across HubSpot and GA4. Inconsistent data between systems was making it impossible to accurately measure campaign ROI and lead quality.
Sedgwick's digital presence spans multiple service lines, audience types, and conversion paths. A tracking approach that worked for one journey broke on another - creating data gaps across different areas of the business.
Data flowing from GA4 into BigQuery and onwards into Snowflake for enterprise BI reporting had to be structured consistently from the point of collection - any inconsistency at the GTM layer created compounding problems downstream.
The analytics implementation had to be maintainable, extensible, and easy to hand over to new team members. A solution that required specialist knowledge to interpret was not acceptable for an enterprise environment.
With multiple stakeholders relying on analytics data across marketing, technology, and business intelligence, data governance and quality assurance had to be built into the process - not added as an afterthought.
The Solution
TagNexus approached this engagement as an architecture problem first and an implementation problem second. Every decision was made with scalability, maintainability, and data quality as the primary constraints.
Before writing a single GTM tag, we conducted a full requirements workshop with Sedgwick's marketing and technology teams to understand what business questions the data needed to answer. From this we built a comprehensive event taxonomy - defining every interaction to be tracked, how it would be named, what parameters it would carry, and how it mapped to business KPIs.
This taxonomy became the foundation for the Solution Design Reference (SDR) - the single source of truth for the entire implementation, shared across analytics, development, and BI teams.
Rather than relying on GTM's built-in click and form triggers - which are fragile and inconsistent across different page types - we designed a structured dataLayer schema that allowed the development team to push precise, clean data at every tracked interaction.
This approach gave us complete control over what was tracked and when, ensured consistency across all user journeys, and made the GTM container significantly simpler to maintain. Any future tracking requirement becomes a simple dataLayer push - no complex GTM trigger logic required.
With the dataLayer architecture in place, GTM configuration became straightforward. Tags were built to listen for specific dataLayer events and forward clean, structured data to GA4. Custom JavaScript variables were used where additional data transformation was needed, and all tags were configured with proper consent checks aligned to Sedgwick's consent management requirements.
The container was documented thoroughly - every tag, trigger, and variable has a clear naming convention, description, and purpose. Any team member can open the container and understand it without needing to ask.
The GA4 property was configured from scratch with custom dimensions aligned to Sedgwick's business metrics, conversion events mapped to actual lead generation outcomes, and funnel configurations reflecting the real user journeys across different service lines.
Cross-domain tracking was configured to ensure consistent user identification across Sedgwick's digital properties, and user-ID tracking was implemented for authenticated users to enable more accurate attribution across sessions.
The existing GTM container was in poor shape - dozens of tags relying on GTM's built-in click variables, element visibility triggers, and auto-event detection. These are inherently fragile: they break when the site's HTML structure changes, fire inconsistently across browsers, and produce data that cannot be fully trusted.
We stripped out all reliance on built-in GTM variables and rebuilt the container entirely around the structured dataLayer. Every tracked interaction now fires from a deliberate dataLayer push - not from GTM trying to infer what happened by inspecting the DOM. The result is a container that is simpler, faster, and produces data that the team can trust completely.
Sedgwick's lead generation forms are a critical conversion point. We implemented two granular form tracking events: form_start - firing as soon as the user focuses on the first field, capturing intent signals before submission - and form_submit - firing on successful form submission, confirming lead creation.
This two-event model gives the marketing team funnel visibility they did not have before: they can now see how many users started a form versus how many completed it, where form abandonment happens, and which traffic sources produce users who engage with forms but don't convert. This data feeds directly into campaign optimisation decisions.
Sedgwick operates two CRM and marketing automation platforms - HubSpot and Act-on - both of which receive form submissions from the website. We implemented reliable tracking and attribution across both platforms, ensuring that form submissions in either system were captured in GA4 with the correct source, medium, and campaign attribution.
The integration required careful coordination with both platforms' form submission events and validation that data flowing into each CRM matched what GA4 was recording - giving the marketing team a single consistent attribution picture across all platforms.
One of the most impactful technical implementations in this engagement was solving a persistent attribution problem: UTM parameters captured on landing were being lost by the time a user submitted a form - particularly for users who navigated between pages, returned in a new session, or interacted with the site across multiple visits before converting.
The solution was a localStorage-based UTM persistence mechanism: when a user arrives on the site via a UTM-tagged URL, those parameters are stored in localStorage. When the user submits any form - regardless of which page they are on or how many sessions have passed - the stored UTM values are programmatically injected into hidden fields on the form before submission, passing the original acquisition source through to both HubSpot and Act-on with the lead record.
This means Sedgwick's marketing team can now reliably attribute every CRM lead back to its original acquisition campaign - even when the user's journey spans multiple sessions or touches multiple pages before converting. The end-to-end attribution chain from paid campaign through to CRM entry is now complete and accurate.
One of the most complex elements of the engagement was ensuring reliable marketing attribution between GA4 and HubSpot. UTM parameters needed to persist correctly from ad click through to form submission and CRM entry, so that campaign ROI could be measured end-to-end.
We validated the full attribution chain - from paid and organic traffic sources through to HubSpot contact creation - ensuring that marketing teams had consistent data across both platforms rather than two conflicting versions of the truth.
GA4's BigQuery export was configured to ensure raw, event-level data was available for Sedgwick's BI team. The event schema and parameter naming conventions from the dataLayer architecture were designed from the start to map cleanly into Snowflake - ensuring that the structured data collected at the browser layer translated reliably into the enterprise data warehouse without transformation issues.
This end-to-end data flow - from website interaction through GTM, GA4, BigQuery, and into Snowflake - is the foundation on which Sedgwick's enterprise reporting is built.
Every element of the implementation was validated before going live. We used GA4 DebugView for real-time event validation, Chrome DevTools for dataLayer inspection at every tracked interaction, and a systematic QA process that tested every event across multiple browsers, devices, and user scenarios.
A full QA sign-off report was produced documenting every test case, its expected outcome, and its actual result - giving Sedgwick's technology team complete confidence in the data quality before the implementation went live.
Implementation Timeline
Stakeholder workshops, business requirements mapping, event taxonomy design, and Solution Design Reference documentation.
Full dataLayer schema design, developer handover documentation, and implementation support during development build.
Full GTM container build, GA4 property configuration, custom dimensions, conversion events, and HubSpot attribution setup.
Systematic QA across all events and user journeys, BigQuery linkage, full documentation handover, and ongoing support.
Ongoing enterprise analytics partnership - initial implementation followed by continuous support, QA validation for new website features, and expansion of tracking coverage as Sedgwick's digital presence evolves.
The Outcome
The impact of the engagement is measured not just in what was implemented, but in how the organisation now uses data.
Working with TagNexus gave us something we didn't have before - confidence in our data. The structured approach to analytics architecture means that when our marketing team looks at a report, they know the numbers are right.
Key Learnings
Every enterprise engagement teaches us something. These are the principles that came out of the Sedgwick project most clearly.
Time spent designing the dataLayer schema and SDR before touching GTM paid dividends throughout the project. A well-designed schema makes every subsequent implementation decision straightforward.
In an enterprise environment, the value of an analytics implementation is only as durable as its documentation. Every tag, trigger, and naming convention needs to be documented for the people who come after.
Knowing that data would flow into BigQuery and Snowflake shaped decisions at the dataLayer level. The naming conventions and event structure were designed to work cleanly in the warehouse - not just in GA4.
With multiple teams depending on the data, a single tracking error compounds quickly. Systematic QA across every event type and user journey is not a nice-to-have - it is a requirement.