Measurable client results

Challenge: The National Flood Insurance Program (NFIP) needed statistical methods to understand the drivers of floodplain development.

Analytical approach: I served as lead statistician on a contract with Booz Allen Hamilton:

• Developed a novel index to quantify the pressure balance between floodplain and non-floodplain land use
• Created the experimental design and initial pilot study statistical model
• Engineered a Python ETL pipeline that consolidated 7,500+ Excel files of geospatial and explanatory variable data into a relational database. Dimension and fact tables let it scale.

Impact: I delivered the statistical methods, a scalable ETL pipeline, and the experimental design within a 4-month part-time contract. The methods were adopted beyond the initial pilot. The Python ETL scripts, relational database, and R analysis files created a repeatable process for planned expansion to other states.

Challenge: The Census Bureau’s data accessibility issues were limiting public engagement. With fragmented analytics and no unified framework for measuring user behavior, product teams lacked the insights needed to prioritize improvements.

Analytical approach: I led user analytics for Census.gov:

• Built a quarterly analytics framework that combined search data, engagement metrics, and customer feedback to guide cross-functional sprint planning
• Ran technical audits with Bing Webmaster Tools and Python to diagnose critical site performance issues
• Wrote Python scripts to extract Census Business Builder user behavior patterns and find improvements

Impact: Technical audits uncovered a critical indexing error affecting 215 pages. I diagnosed and resolved it, driving a 68.2% increase in new users within one month.

Census Business Builder analysis revealed where users struggled and what functionality they valued most. Translating these insights into development requirements, I identified $1.56M in potential annual savings through software consolidation.

Quarterly executive presentations shifted how teams prioritized product strategy. I also trained the product teams on the analytics framework.

Challenge: Stanley Consultants stored applicant and employee data in separate systems, which blocked analysis across hiring and retention. HR teams spent 40+ hours monthly on manual compliance reporting. Leadership lacked visibility into hiring pipeline effectiveness and retention trends.

Analytical approach: I worked with the systems analyst and HR team to design a data integration:

• Shadowed the HR manager to map workflows and analyzed both database structures to find integration points
• Identified unique applicant IDs as the bridge between disconnected datasets
• Designed a custom field that tracks each person from candidate to employee
• Partnered with the VP of Operations and Chief of HR to standardize retention and turnover methods
• Established access controls and security protocols for sensitive workforce data

Impact: The disconnected systems hid conversion rates, time-to-hire, and retention patterns. My custom field created full lifecycle tracking without an expensive ATS. Streamlined affirmative action reporting cut HR administrative time by 40 hours per month.

The unified data showed the root cause of a long-standing disagreement between the VP of Operations and the Chief of HR on retention calculations. I set a standard reporting method that resolved the conflict and produced the company’s first workforce analytics.

Challenge: Mosaic Co, a Fortune 500 fertilizer producer, needed to optimize its Azure data warehouse. Siloed departmental systems (operations, HR, safety, finance) duplicated data and blocked cross-system integration. That raised the engineering burden and limited enterprise-wide insight.

Analytical approach: During a 6-month consulting engagement with Bii Consulting, I analyzed the data warehouse:

• Queried the siloed systems in SQL to map data flows and document redundancies
• Identified cross-linking opportunities that let departments share data without duplication
• Inventoried the Power BI reporting library to find and remove redundant reports

Impact: The analysis found roughly 3x data duplication across departmental silos. I created Mosaic’s first centralized enterprise database dictionary, which gave the data engineering team a full map of their data. That documentation, plus testing in Azure and Power BI, identified consolidation opportunities for the organization’s optimization roadmap.

Challenge: Healthy Hildegard scaled to 10,000 monthly visitors across two countries. Three barriers held back revenue: slow international site speeds, coordinated payment gateway attacks, and high cart abandonment.

Analytical approach: I set up analytics across site performance, security, and conversion:

• Analyzed page load times and geographic patterns to identify international customer friction
• Monitored payment gateway activity to detect and prevent fraudulent attack patterns
• Ran drip campaign A/B tests on email timing, content, and cart recovery messaging

Impact: Performance data showed that geographic distance caused the slow load times driving international customers away. I set up the Cloudflare CDN and migrated to SiteGround hosting, which raised conversion rates for international customers.

Security analytics detected coordinated bot attacks targeting the payment gateway. I configured automated IP blacklisting rules and optimized the Authorize.net API integration to prevent fraudulent transactions and protect customer data.

Marketing analytics shaped email automation for cart recovery and order fulfillment. It cut manual processing time and recovered abandoned revenue.

Challenge: Lat-Lon needed more qualified leads for its IoT asset tracking devices in a competitive B2B market. With a limited marketing budget, every ad dollar had to show a clear ROI.

Analytical approach: I designed and ran an A/B testing framework for LinkedIn ads and website content:

• Tested ad copy (technical vs. business value messaging), visual elements, and call-to-action placement
• Analyzed audience segmentation by industry, company size, and job titles
• Evaluated platform performance and engagement patterns across channels
• Optimized website content and search strategy to improve discoverability

Impact: Ads about data-driven decision-making outperformed feature-focused ads on lead quality.

Web content analysis showed which case studies and keywords resonated with target audiences. I used those results to rework the ad strategy and website architecture. Ad click-through rates and search visibility rose, cost per acquisition fell, and the sales team received higher-quality leads.