Episode #97: Predicting Product Adoption with UX Research and Lawton Pybus

A polished product and enthusiastic usability feedback do not guarantee that people will return six months later. In Episode #97, human factors researcher Lawton Pybus explains how product teams can study likely adoption without asking users to predict their own future behavior. The conversation connects established research models with startup constraints, business decisions, and the growing use of AI in UX research.

What you'll learn

• Why perceived ease of use and perceived usefulness can indicate later product adoption

• How a short two-item measure can create a baseline for an early product or prototype

• When fast research is appropriate and when a decision deserves greater depth

• Why researchers need to translate findings into business direction

• How to use AI tools without overlooking bias, limitations, or foundational research methods

Measure the conditions behind adoption

Directly asking whether someone will use a product in six months produces an uncertain answer. As Lawton notes, people are not well equipped to forecast their own future behavior. Human factors research offers another route through the technology acceptance model. It focuses on two perceptions: whether a product seems easy to use and whether it seems useful. Together, those measures can provide evidence associated with later use.

Lawton describes a condensed instrument used by UX professionals with two agreement questions. One asks whether the product was easy to use. The other asks whether it met the participant's requirements. An early-stage team can ask these after a usability session involving a prototype or demo. The results create a baseline, then give the team something consistent to measure as the interface changes.

This does not turn adoption into a certainty. It improves the question. Instead of relying on a participant's promise about future use, the team measures two conditions that research has connected with acceptance. That distinction gives founders and product teams a more disciplined way to evaluate progress toward a product people can understand and find valuable.

Match research depth to the decision

Lawton does not treat scrappiness as a virtue by itself. Research should reduce uncertainty and risk by giving a team good information. The appropriate method depends on the realities of the decision. If a stakeholder needs an answer by Friday, the researcher has to choose the best credible approach available within that limit. If a six-month study will shape strategy for several years, speed is less important than planning carefully and drawing out richer evidence.

The same principle applies as a company builds a UX function. Founders and product managers may initially conduct research themselves. Specialization can add value as the organization grows, but hiring an academically accomplished researcher is not enough. Lawton warns that researchers from academic settings can struggle to connect rigorous findings with business needs. In a company, stakeholders often need a recommendation, not only a list of study limitations.

Strong researchers therefore explain what the evidence means for stakeholder goals and key performance indicators. They acknowledge uncertainty while still helping the team decide what to do next. Rigor matters, but it has to serve the decision rather than delay it indefinitely.

Use AI with openness and informed skepticism

Lawton sees potential for large language models to analyze sources that were previously difficult to examine at scale, including app reviews, customer reviews, support tickets, and other batches of unstructured text. These sources were not created as formal research data, but suitable prompting may help teams extract relatively reliable information from them.

His advice combines curiosity with technical understanding. Researchers should remain open to what a new tool can do, while learning enough about how it works to judge where it may perform well or fail. An uncritical approach can produce a wrong answer faster, especially when models reflect biases in their training data or are treated as substitutes for specialist work.

New tools also do not erase older foundations. Lawton points to mental models, quantitative ways to measure how those models change, and skill transfer from one system to another as areas UX teams can study more rigorously. The technology changes, but attention, memory, decision-making, and the purpose behind a user's actions remain central.

Listen to Episode #97

This conversation offers a practical way to connect adoption research, business judgment, and responsible AI use. Listen to Episode #97 with Lawton Pybus for the full discussion.