15% Increase In Revenue Through A/B Testing Program and AI workflows.

Data Governance:

• Coordinated with developers and teams to architect a robust data pipeline for Segment, GA4, ad platforms, and Klaviyo.
• Achieved accurate, reliable data across platforms to enable actionable analysis.

Experimentation:

• Conducted UX/UI assessments with heatmap analysis to identify pain points and develop hypotheses.
• Configured and ran A/B tests using LaunchDarkly, improving clickthrough rates and conversion rates by 10-15%.
• Introduced a new feature on marketing pages, now in full development, significantly enhancing user engagement.

Growth Model & Dashboard:

• Developed multiple growth model scenarios to align team responsibilities and foster collaboration.
• Built dashboards for real-time tracking, enabling teams to monitor progress against targets and identify improvement opportunities.

Notarize faced several challenges that hindered their growth:

• Disjointed data pipelines led to inaccuracies across GA4, ad platforms, and Klaviyo.
• A lack of systematic experimentation slowed progress on user engagement and conversion optimization.
• Teams needed better visibility into growth metrics and clearer alignment on responsibilities.

Planning:

• Collaborated with developers, designers, and stakeholders to define data architecture and establish event tracking requirements.
• Scoped experiments with UX/UI assessments, developing hypotheses for A/B testing.
• Aligned teams on growth model responsibilities and designed dashboards for performance tracking.

Execution:

• Implemented a data governance plan for accurate, consistent data across all platforms.
• Configured LaunchDarkly for A/B testing, running multiple experiments to validate hypotheses.
• Delivered dashboards to enable teams to gauge progress and make data-driven decisions.

Increasing Engagement and Driving Conversion Through Targeted Experiments

Objective:

Improve click-through rates (CTR) and conversion rates by 10-15% through targeted experimentation. Enhance data-driven decision-making with actionable insights while fostering team alignment through clear roles and real-time reporting tools.

Experiment 1: Aligning Landing Pages with Ad Messaging to Boost Conversion Rates

Problem Statement:
Multiple ads directed users to the same generic landing page, potentially creating a disconnect between user expectations and page content. We hypothesized that aligning landing page content with specific ad messaging would provide a more cohesive user experience, increase trust, and encourage document uploads — a key conversion step.

Research:
Prior to the experiment, we observed that users from various ads were funneled to identical landing pages. This uniform approach likely reduced relevance and trust, impacting conversion rates. The metrics analyzed included:

• Click-through rates (CTR) on CTAs
• Document uploads
• Purchase behavior

Hypothesis:
If landing pages are tailored to match ad messaging, users will experience a more consistent journey, leading to higher document uploads and increased conversion rates.

Design & Implementation:

• Type: A/B test comparing the default generic landing page with newly designed, ad-specific landing pages.
• Duration: 2 weeks
• Metrics Tracked: CTA click rates, document uploads, purchase conversions
• Sample: Three distinct landing pages, each aligned with specific ad messaging

Results:

• 2 out of 3 new landing pages outperformed the control.
• CTA click rates increased by 5-10%, leading to a 10% improvement in document uploads.
• Purchase rates showed a minimal increase, indicating increased engagement but not a significant lift in final conversions.
  
  

Conclusion
The experiment validated the hypothesis that aligning landing pages with ad messaging improves user engagement and document uploads. Next steps include:

‍

• Refining the most successful landing pages
• Extending the test duration to further assess purchase conversion impact
• Applying the tailored approach to additional campaigns

Experiment 2: Increase Conversion of "Uploaded Document" by "Making Requirements" More Visible"

Problem Statement:
A heatmap analysis of the upload page revealed that users rarely interacted with the hidden requirement list. This likely led to confusion and higher churn rates during the upload process.

‍Research:
The initial hypothesis stemmed from user behavior analysis on a similar page on Notarize. Heatmap data confirmed that users often clicked on the text labeled “#1,” likely mistaking it for the upload CTA, while ignoring the hidden requirements list.

Hypothesis:
By making the requirement list immediately visible and emphasizing the 'Upload your documents' CTA, we expected to increase the session-to-upload and session-to-user-creation rates, minimizing churn.

Design & Implementation:

• Type: Multivariate A/B test with three variants:
  • Control: Existing upload page with hidden requirements
  • Variant #1: Upload page with visible requirements
  • Variant #2: Same as Variant #1 but with different wording
• Duration: 2 weeks
• Metrics Tracked:
  • Session-to-user conversion rate
  • Upload CTA click rates
  • Account creation rates
  • Purchase conversion rates

Results:

• Primary Data:
  • Control: 27.33% session-to-user conversion rate
  • Variant #1: 30.24% (+10.64% over control)
  • Variant #2: 29.9% (+9.4% over control)
• Secondary Data:
  • Due to Mixpanel access limitations, data for account creation and purchase rates was unavailable.

‍

‍Default Experience – User is displayed the upload page with the requirements details hidden under a dropdown text button. Only by clicking the #2, #3 and #4 text buttons the details are revealed.

‍

‍Variant #1 – User is displayed the upload page with the open list, informing the user about the requirements 

‍

Variant #2 – Same as Variant #1 but with different wording

Conclusion:
The experiment confirmed the hypothesis: Making the requirements visible significantly improved session-to-user conversion rates. Next steps:

• Further refine the messaging and structure based on user feedback
• Extend the test to gather more data on purchase conversions
• Explore additional iterations to address remaining friction points