Request a review

Selected work

Measurement problems solved across analytics, consent, attribution, and CRM.

Atem Analytics supports agencies and in-house teams across complex measurement environments. Many engagements are completed white label, so client names and identifying details are withheld. The case studies below show the type of problems I am brought in to solve, how I approach the work, and what was delivered.

Discuss a similar projectRequest a tracking review
Case 01Direct-to-consumer ecommerceConsent measurement

Separating a real consent improvement from a reporting illusion.

The situation

A direct-to-consumer eyewear brand changed from a single California-style consent experience across all US traffic to a state-tiered configuration.

Reported opt-in performance increased immediately. On the surface, the new number appeared to show that significantly more visitors were actively accepting tracking. That was only part of the story.

What we found

The change in the headline rate was being influenced by two separate factors: more total website traffic, and a change in how visitors were classified under the new regional configuration.

Some visitors were now entering a more permissive default state. That improved the reported consent metric, but it did not mean every additional visitor had actively opted in. Reporting the full increase as a change in user behavior would have overstated the result.

What we did

We separated the performance change into its actual components. The analysis reviewed:

  • 01Traffic volume before and after the change
  • 02Consent rates by region and configuration
  • 03Active consent compared with default classification
  • 04Tag behavior across consent states
  • 05Effect of the revised experience on analytics and advertising measurement

Outcome

The final analysis separated approximately 25% traffic growth from an approximately 19% improvement in the measured consent rate.

It identified the classification effect so the marketing, privacy, and legal teams could understand what the headline number did and did not represent. The state-tiered implementation was validated against the required tag behavior, and the report was delivered in HTML, PDF, and editable Word formats.

Services involved

Consent measurement analysisRegional consent validationTag-blocking QAStakeholder reportingTechnical documentation
Case 02Shopify ecommerceGA4, GTM, Search Console

Diagnosing a reported traffic decline that started outside the analytics platform.

The situation

A Shopify retailer believed its organic search performance had been penalized. GA4 sessions had declined, leads were down, and the existing reporting did not provide a clear explanation.

The team needed to know whether the problem was search visibility, user behavior, measurement, or a combination of all three.

What we found

Google Search Console showed that search impressions remained relatively stable while click-through rate declined sharply during a specific mid-June period. That pattern was more consistent with changes to the search results page and increased SERP competition than a broad ranking loss.

The audit also identified several unrelated measurement issues that were making the overall performance picture less reliable.

Measurement findings

  • 01An additional GA4 measurement ID loading alongside the intended property
  • 02Data retention set to two months, restricting long-term analysis in Explorations
  • 03A high-value quote journey moving visitors to a third-party environment without reliable submission tracking
  • 04Configuration and search findings being interpreted as one problem despite different causes

What we did

We reviewed the website, Google Tag Manager, GA4, Search Console, and the third-party quote journey. The final report separated:

  • 01Organic search visibility
  • 02Search-result click-through behavior
  • 03On-site analytics configuration
  • 04Lead-tracking gaps
  • 05Data-retention limitations
  • 06Recommended implementation priorities

Outcome

The team received a clearer explanation of the apparent traffic decline without incorrectly treating it as a broad ranking penalty.

  • 01Additional GA4 property issue documented for removal
  • 02Data-retention limitation identified and corrected for future analysis
  • 03Untracked quote journey prioritized as a significant lead-measurement gap

Services involved

GA4 auditGoogle Tag Manager auditSearch Console analysisCross-domain journey reviewLead-tracking strategyImplementation planning
Case 03Specialized healthcareHubSpot workflows & integration

Building a scalable HubSpot onboarding system around an external signup process.

The situation

A specialized healthcare organization wanted onboarding emails to reflect the partner network through which each new client registered.

The registration process occurred outside HubSpot. New accounts were created in an external system and later synchronized to the CRM through a backend integration. The partner identifier therefore had to survive the complete signup and synchronization process before HubSpot could use it for personalization.

The challenge

The existing onboarding workflow was already active. Changing it directly would have created unnecessary risk for contacts currently moving through the sequence.

The solution also needed to support additional partner networks without creating a separate workflow for every landing page.

What we did

We cloned the existing workflow and built a controlled routing structure. The revised setup used:

  • 01One centralized HubSpot contact property
  • 02Separate paths for partner and standard registrations
  • 03HubSpot smart content for partner-specific messaging
  • 04A shared onboarding sequence after the personalized welcome step
  • 05A staged migration plan for contacts already enrolled

Post-launch monitoring

The framework was later extended to six partner networks. During monitoring, two referral lists remained at zero even though the HubSpot workflow logic was operating correctly.

We traced the issue upstream to the backend property synchronization, documented the affected cases, and escalated the defect to the development team. This prevented unnecessary changes to workflow logic that was already working as intended.

Outcome

  • 01One-partner pilot expanded into a six-network onboarding framework
  • 02A single centralized workflow replaced duplicated automations
  • 03Existing contacts completed their original journeys without interruption
  • 04Post-launch QA isolated an integration issue before it was treated as a workflow problem
  • 05Structure extends to new partners by adding property and content rules

Services involved

HubSpot workflow architectureCRM property strategyEmail personalizationIntegration requirementsMigration planningPost-launch monitoringDeveloper coordination

Our process

Find the real failure point before changing the system.

  • 01Start with the business problem. Clarify what stopped making sense, which decision is affected, and what the team expected the system to do.
  • 02Review the full journey. The issue may sit in the website, tag manager, analytics property, advertising account, CRM, consent platform, backend integration, or reporting layer.
  • 03Separate related symptoms. A decline in reported performance can have several causes. I avoid forcing every symptom into one explanation.
  • 04Protect what is already working. Live workflows and tracking are tested and staged carefully rather than changed without a rollback or migration plan.
  • 05Validate after launch. Implementation is followed by QA, monitoring, and documentation so upstream or downstream issues are identified early.
  • 06Make the findings usable. Reports explain the business impact, technical cause, recommended priority, dependencies, and next action. White-label and editable versions are available for agency partners.

Other engagement types

The selected cases represent only part of the work.

Atem Analytics also supports:

Bring me the unclear problem

Have a client environment where the numbers no longer add up?

Send the website, platforms involved, and a short explanation of what appears to be wrong. I'll help determine whether the next step should be an external review, account-level audit, implementation project, integration investigation, or ongoing support engagement.

Discuss a project Request a tracking review