One full case study with a measured result, four real Control-vs-Variation tests that show how I write a hypothesis and design a variation, and a handful of other builds along the way.
A lead-capture popup offering 20% off had a gap between "submit your email" and "actually use the code." Closing that gap was the whole test.
Once a user submitted their email for a 20%-off code, the popup just displayed the code. There was no visibility into what the discounted price actually was, and nothing that made using the code feel time-sensitive — so a lot of codes went unused.
If users see their real, discounted price immediately — with a visible countdown on the offer — they're more likely to complete the purchase in that session instead of abandoning with an unredeemed code.
Rebuilt the post-submit state of the popup: auto-scrolled the user down to the real price with the discount already applied, added a live countdown to the offer window, and triggered a confetti animation the moment the coupon successfully applied — copy, motion, and a small reward stacked together.
Conversion rate from popup email-submit to completed purchase, tracked in Optimizely and cross-checked in GA4 within the same session window.
on the tested flow, shipped as the new default experience.
These show the hypothesis-writing and variation design side of the job — the step before a test goes live. Happy to walk through launch data and reporting for any of these in conversation.
Hypothesis: Showing an itemized price breakdown before the final action reduces hesitation at checkout.
What changed: Replaced a bare terms-and-conditions checkbox with an Order Summary card — base price, applied coupon, total, and a money-back guarantee badge — directly above the Continue button.
Hypothesis: Reducing the comparison detail shown by default lowers decision effort and increases enroll clicks.
What changed: Moved the full feature-comparison table behind a "View Package Details" link under each tier, so users can choose straight from the cards or dig into specifics on their own terms.
Hypothesis: The visual position and weight of "Claim Promo" versus "Course Only" changes which path users default to.
What changed: Swapped the column order and button emphasis so the promo path sits in the position and style most likely to be chosen.
Hypothesis: Inconsistent visual weight across entry options adds decision friction and suppresses click-through.
What changed: Replaced four unevenly styled tiles with four uniform, high-contrast pill buttons and a simplified ratings line, so every option reads with equal confidence.
Additional experiments and UX builds from Aceable and NFP.
A second lead-capture popup variant testing discount framing and urgency copy.
Built to replace a plain dropdown for state-specific course eligibility.
Dual-path search — by person or by office — built for NFP's national site.
Built during an enterprise CMS migration for a 400+ location brand.
I keep a live testing backlog and full reporting for every experiment I run. Ask me about any test above.