How a small-business website project produced useful evidence through focused questions, mixed methods, and honest limitations
A lack of research funding does not eliminate the need for evidence. It changes the way the research must be planned.
I saw that firsthand while leading UX research for the redesign of the Monarch Services & Construction website. The small business had no dedicated research budget. The existing website had been launched in 2012, contained only three pages, was not mobile-friendly, and provided too little information for homeowners trying to decide whether they could trust the company.
The easy response would have been to rely on stakeholder opinions and redesign the site based on assumptions. Instead, I built a mixed-method research program using tools already available to me, uncompensated participants recruited through Facebook and email, and analysis in Google Forms, Optimal Workshop, R, and Excel.
The result was not a perfect substitute for a funded study. It was still meaningful research that answered real business and design questions.
When resources are limited, every research activity has to earn its place. I began by identifying the decisions the business needed to make:
Those questions determined the research sequence. I did not select methods because they were familiar or impressive. I selected them because each one reduced a different kind of uncertainty.
The work began with a UX audit of the existing site. It identified excessive vertical scrolling, difficult-to-find navigation, readability problems, limited company information, and missing content, such as services, testimonials, and project examples.
Next, I conducted ten user interviews. Eight of the ten participants said they looked online for home-improvement contractors. Trust signals repeatedly surfaced in their responses: referrals, testimonials, BBB ratings, a professional-looking website, and before-and-after photos. None of the ten could find Monarch through a search engine, revealing a business problem that extended beyond interface design.
A quantitative survey then broadened the picture. I used Google Forms and recruited through a Facebook post and an email link, with no participant incentives. Thirty homeowners responded. The findings showed that 78% hired professionals because they lacked the required skills or knowledge, 66% had hired a contractor for renovation work, and 63% wanted accessibility features in future renovations.
The survey also supported more detailed comparisons. Because the age-group samples were small, I used multiple statistical tests—including chi-square, Fisher’s exact, Kruskal–Wallis, and ANOVA—rather than treating descriptive differences as proof. Most behaviors were similar across age groups. The primary exception involved where people looked for home-improvement information: younger groups relied more on online sources, while the oldest group relied more on personal experience and friends.
First-click testing evaluated the proposed navigation before detailed design work. Participants were recruited through Facebook and email and completed the study without incentives. Task success ranged from 80% to 100%, and average first-click times ranged from 2.64 to 9.82 seconds. A one-way ANOVA found no significant performance differences across age groups, F(2,21) = 0.28, p = 0.76. The navigation worked consistently, while a few labels—especially those connected to company experience and reputation—needed refinement.
Finally, Conjoint, MaxDiff, and TURF analyses helped prioritize renovation options and marketing content. Homeowners consistently valued practical features such as energy-efficient appliances, smart storage, under-cabinet lighting, quartz countertops, and semi-custom cabinetry. TURF analysis identified a feature combination that reached 100% of participants.
The studies produced a research-backed roadmap rather than a collection of interesting observations. Recommendations included:
The research connected website decisions to customer needs and business outcomes. It explained why homeowners selected contractors, what evidence they needed before making contact, which services and features mattered, and where the existing experience created friction.
The constraints mattered. Participants were self-recruited, uncompensated, and drawn from my available Facebook and email networks. The samples were small and were not representative of every homeowner or every potential Monarch customer. Those limitations affect how confidently the findings can be generalized.
I treated the results accordingly. I did not present small differences as universal truths. I used statistical tests where they were appropriate, reported when differences were not significant, and looked for patterns that appeared across more than one method.
That triangulation mattered. Interviews showed that trust influenced contractor selection. Survey results reinforced the importance of referrals and professional expertise. The audit identified missing credibility content. Search tasks showed that potential customers could not find the business. Each method added a different piece of evidence.
The value of this work did not come from an expensive platform or a large sample. It came from a disciplined process:
A research budget would have made broader recruitment, participant compensation, and specialized tools possible. Those things would have strengthened the work. But waiting for ideal conditions would have produced no evidence at all.
Meaningful UX research is not defined by how much money was spent. It is defined by whether the research asks a useful question, applies an appropriate method, acknowledges its limitations, and helps people make a better decision.
Draft source material: Monarch Services Case Study; User Interview Results; Quantitative Survey Results; First-Click Test Results.
Open to contract UX research opportunities.
Email: theresaw@columbus.rr.com
LinkedIn: theresa-wilkinson