UX Researcher · Mixed-methods & experimentation

I find the why and prove the what.

I build research practices from the ground up and turn messy human behavior into decisions teams can measure. Qualitative depth, quantitative rigor, and findings that the team actually uses.

5+ years in research Interviews · Surveys · Usability A/B testing & analytics B2B SaaS · DTC · Insurance

Selected work

From building an entire research function to launching quick, low-budget surveys. Figures are rounded, and some details are generalized to protect proprietary work.

Research ops · Practice building

Building the research function from scratch

I joined a company with no formal UX research and built one the team now reaches for by default.

0 → 1

From no research function to a practice that reaches the CMO, CFO, and company President, and that teammates bring problems to without being asked

Context

When I joined Independence Pet Group, there was no formal UX research function. Decisions were made without a shared way to learn before a launch or validate after one.

Where I started

I led with analytics reporting and, more importantly, with people. Before pushing any process, I spent time getting to know teams and earning trust, because a research practice only sticks if the people around it want it there.

What I built

I prioritized the two gaps that were costing the most: teams shipped without learning first, and whatever we did learn vanished after each project. So I onboarded the company's first user-testing tool and introduced a habit that hadn't existed: pre-launch ideation through usability and customer research, so teams could learn before they built rather than after. I created a research intake process that grew into a searchable dashboard where the whole team can track and dig through live and past studies. The experimentation side of the practice grew into a program of its own, covered in its own case below.

Research opsUsability testingInterviewsToolingStakeholder trust

Going broad

I ran brand-focused survey and interview research across several of the portfolio's top-performing brands, surfacing dozens of insights that led to multiple successful design changes, including the plan repackaging detailed in the next case.

Impact

By being a reliable partner and always offering a hand, trust and understanding of the research process compounded. I reached a point where people propose tests on their own and come straight to me with the questions they want answered. I present findings regularly to the CMO, CFO, and President, and several have circulated independently among the executive team because of their strategic relevance.

What I learned

Building a research function from the ground up taught me that kindness and always offering a helping hand are fundamental to building a practice that non-researchers are invested in. The methods matter, but the relationships are what make the methods count.

Mixed-methods · Conversion

Why pet owners weren't converting

A two-week study that reframed the product from forced customization to need-based plans.

8 figures ↑

Projected annual increase in channel revenue from the redesign · shipped, and validated by an A/B test before launch

Context

This is a direct-to-consumer pet insurance product, and the people shopping for it are pet owners trying to do right by their animals. Many were showing real interest in the product but leaving the online sign-up experience before enrolling in a plan, and the team had no solid evidence to explain why.

The study

The question had two halves: how widespread the drop-off was, and why it was happening. So I matched a method to each, running a two-week mixed-methods study across both current customers and pet owners who had looked into insurance, but never enrolled their pet. A survey of roughly 650 respondents sized the patterns at scale; 12 in-depth interviews surfaced the reasoning behind them.

Survey (n≈650)Interviews (n=12)Existing + non-convertedTriangulation

What I found

The study surfaced many insights, two of which carried the outcome of this study. First, a mental-model mismatch: people assumed pet insurance worked like human health insurance, which made the plan structure hard to parse. Second, the old flow forced people to customize their plan by default, and for someone already overwhelmed and facing an unfamiliar product, that led many of them to give up or go to another provider with a more intuitive plan selection experience.

Recommendation & impact

I recommended moving from forced customization to need-based, pre-packaged plans that meet people where they are. The redesign shipped, boosted relative conversion rates by double digits, and is projected to add an eight-figure increase to annual channel revenue. We validated it with an A/B test before rollout, detailed in the next case.

Experimentation · Program building

An experimentation program from zero

Built the company's first A/B testing capability, the cross-functional group around it, and a track record of wins, including one valuable decision not to launch.

+16.5%

Conversion lift

+17.4%

Channel revenue increase

Cumulative, over ~7 months · cadence scaled from 0 to 6 tests per quarter

Context

No experimentation practice existed. Design and marketing calls were largely made on instinct, with no structured way to learn from them.

What I built

Running a few one-off tests would not have changed how the company made decisions, so I built a repeatable system rather than a pile of experiments. The whole loop, from scratch: test intake, prioritization, design, analysis, and readouts. I founded a cross-functional working group across UX, Creative, and SEO that meets for biweekly ideation, and I introduced retrospectives to review what each round of tests teaches us.

A/B testingPrioritizationStatistical significanceWorking group

Selected wins

+17% / +20%

Conversion / channel revenue per user · 99% conf

Validated the plan-repackaging redesign from the previous case before launch: conversion up 17% and channel revenue per user up 20%.

+16%

Leads · 99% conf · <15 min build

Changed one button from "Get a Quote" to "See Pricing," based on research showing pricing is the first thing pet-insurance shoppers look for. It has funneled tens of thousands more leads since.

+19%

Wellness add-on sales · 99% conf

A personalized experience targeted at new-puppy owners lifted wellness add-on attach rates.

−20%

Conversion in test · not launched

We expected a win, and tested it anyway before committing. It cut conversion by a fifth, so we killed it before launch. Catching a costly idea early is worth as much as any lift.

Impact

Ideas now get tested before they go live. We've scaled testing from zero to six tests per quarter, and over a specific seven-month period of targeted optimizations, cumulative testing drove a ~16.5% lift in relative conversion rate and a ~17.4% increase in relative channel revenue.

Survey · Marketing efficiency

A $500 study that moved six figures

A lean survey that showed a signup incentive wasn't landing or being seen, and redirected the spend to higher-performing channels.

<$500

Total research cost

6 figures

Budget redirected to higher-performing channels

Context

The company was spending on a gift-card incentive to drive signups, with little customer-side evidence that the offer was actually working.

What I did

This question didn't call for an expensive study. It needed a fast, clear read on two things: whether people knew about the offer, and whether it moved them at all. So I kept it deliberately lean, a post-signup survey under $500 all in, asking new customers whether they had been aware of the offer before signing up and how much it influenced their decision, with branching follow-ups to drill into the underlying motivation.

Survey designBranching logicLean / scrappy

What I found

The data pointed to two issues at once: the incentive wasn't compelling to most of the audience, and most weren't even aware of it before they signed up. It was neither landing as an offer nor being seen as a piece of marketing.

Impact

I took the findings straight to our CMO and made the case for shifting the spend. The recommendation landed: on the strength of a study that cost under $500, a six-figure budget was redirected toward higher-performing channels.

How I work

Question first

I scope what a decision actually needs before choosing a method, and I'm candid about what each method can and can't tell you.

Scrappy to deep

A $500 survey can be as decisive as a months-long study when it's scoped well. I match the rigor to the stakes.

Insight that travels

Research only counts when it's used. I turn complexity into a clear, prioritized story that non-researchers can act on.

Mixed-methods researchIn-depth interviewsUsability testing (moderated & unmoderated)Survey designA/B testingInformation architecture

What drives the work

Keep human beings at the forefront, and drive change that makes the world a better place for everyone.

About

Anthropologist by training, researcher by trade.

My path into research started with a degree in Anthropology at UC Berkeley, and my studies shaped how I think: watch what people do, not just what they say, and take context seriously. I've spent the years since turning that lens on products, from analyzing hundreds of B2B SaaS experiments a week at an experimentation firm, to building a research practice from the ground up in-house.

I care most about the moment research stops being a deck and starts changing a decision, and about keeping real people at the center of the products built around them.

2022 — NowIndependence Pet Group — UX Researcher

2021 — 2022DoWhatWorks — UX / Growth Analyst

2020 — 2021Berkeley Disability Lab — Lead Design Researcher

2018 — 2020UC Berkeley — B.A. Anthropology, 3.9 GPA

When I'm not knee-deep in research, you'll find me hiking and biking around Los Angeles, hosting dinner parties for friends and family, or on the couch with Figgy, my rescue Bulldog.

Get in touch

Let's talk research. Say hello →

everett514@gmail.com (818) 450-6743 LinkedIn ↗