Mary L. Fraser

Research is about RelationshipsI help teams and companies develop design startegies by executing research aimed at users and business KPIs.

I am a fintech/software researcher, with explorations in AI, infosec, and trust. I have done cloud, enterprise, and B2C with great success.Twice, I have built zero-to-one teams of UX Researchers and led teams in design research, CX, and UX. Being a player/coach keeps my skills sharply attuned to the needs of my organization and users while allowing me to stay as strategic as I can.Exploring the messy questions around how business can understand users is why I love research.

Wealth is not a number; it is a feeling. Freedom and Trust in relationships over financial transactions are valued.X Bank was losing its wealthiest clients, including the younger generations of familial wealth.How do wealthy segments of each generation prefer to relate to their bank.61 In-Depth Interviews determined segmentation across generations of wealth, as well as five pathways to wealth.1. A hybrid roboadvising tool was implemented as a synthesized tool (specific to basis points)
2. Better relationship skill training for advisors and clients was implemented.An 14% increase in client retention was the overall result.

UXR is not Obsolete, Not even close

My thoughts are that a huge potential opportunity is offered by using AI to enhance rather than replace human-led research practices.My research partner and I brainstormed ways in which AI could be used to run a study in the following spaces: fintech, healthtech, education, and trust/safety. We asked ChatGPT5 the same question.We then asked other UXRs in each domain (Na=20, na1=5x4) to identify which hypothesis might be more achievable or measurable and why (because AI spews nonsense sometimes - see hallucinations)We pitched three alternatives against each other1. Plans developed by human researchers alone
2. Plans developed by ChaptGPT5 alone
3. Plans developed by human researchers in collaboration with ChatGPT5The result was overwhelming. (83%) preferred the plan developed as a hybrid between AI and Humans.These numerical results reinforce the sense we have heard from many teams that human+AI research beats AI alone.