User Experience Research

Understanding the diverse perspectives and human experience of people who use a given product is at the heart of user experience (UX) research. In my work, I seek to understand the human experience at scale to deliver key insights that allow product teams to improve or expand a given product or create new product experiences.

Some of my research is foundational — unpacking the underlying landscape for a product to reach product market fit. Other times, my research is tactical in trying to identify top pain points or the most desired features for a product experience. To get generalizable insights at scale, I leverage survey research methodology to randomly sample key populations of interest and generate reliable insights for product teams to action upon. I leverage these analyses to craft product narratives and formulate strategic direction across product teams and across orgs. Below, I expand on some of the methods I use in my work.

UXR METHODS

Research Design

Questionnaire Design: Appropriate questionnaire design is essential for generating unbiased UX insights. Questionnaire design must consider careful question construction, item response scales and ordering, the order of the questions, and cognitive burden, among other factors. Additionally, methods such as branching, randomized question selection, randomized survey prompts, and MaxDiff designs expand the realm of possibility in UX surveys. While most of my work is quantitative, my skills in questionnaire design translate to writing semi-structured interviews where qualitative UXR is needed.

Sample Design - Databases, SQL, and Python: Quantitative UXR requires precise sampling designs for key user groups. I regularly develop SQL and Python based data pipelines for sampling key populations. For example, we might want to sample people in a given country who fit a niche user profile based on myriad product-specific factors and other key dimensions of interest to stakeholders. Pipelines typically involve aggregations, complex joins, and multiple subqueries.

International Survey Research: Although UXR often occurs on US-populations, product growth frequently requires international insights and a foundational understanding of new markets. To this end, I have run international surveys in over 20 countries to deliver impactful research.

Research Analysis

Post-Stratification Weighting: Once the survey is run, post-stratification weighting can help us better understand the unbiased state of the underlying population compared to the raw results of the survey. This is especially important for scenarios where you oversample user populations whose behavior systematically differs from the typical user.

Computational Data Analysis: Depending on the project, I also leverage SQL and Python for computational data analysis — providing insights on all users in a population, or conversely, focusing on the subset of users for which we also have survey data. We can, for example, use machine learning or multivariate modeling to yield predictive insights about user behavior.

Visualization: To tell a compelling user journey, visualization of quantitative survey data is key. Most often, a series of bivariate bar-charts with confidence intervals — for example, top pain points or top features by user group — makes a compelling case to give a team direction regarding what product features to develop. Even where more robust methods, such as multivariate modeling or machine learning are used, I prefer to visualize those results so the takeaway can be easily discerned for stakeholders. I make most of my visualizations in R, generalized examples of which are below.

Visualization

As mentioned above, a number of potential types of UX analyses output exist. I have included some visualization archetypes that are possible using generalized, hypothetical UXR scenarios. All below figures are representative of the my work, however, all data have been simulated and all specific details removed to preserve confidentiality.

International Product Feature Sentiment: An example UX scenario I have worked on previously is evaluating user sentiment for a key product feature. Once we know their baseline sentiment, I follow up to know why they felt that way. In this way, we know better where to iterate to improve product favorability.

Top Pain Points for a Key User Group of Interest: I have also sought to better understand the top pain points for a given strategic user group so that product teams might develop features to grow and retain one or more strategic groups of interest. In this way, we can resolve user friction and improve the product experience.

Barriers to Consideration, Barriers to Use: Similar to top pain points research, at times we would like to better understand what prevents a key user group from considering a product or product feature. What are their top barriers to consideration? If they would consider using a product but do not, what are their barriers to use?

Top Use Cases for a Feature or Top Solutions to a Product Issue: Sometimes we are already aware of top pain points and barriers but we don't necessarily know how best to solve the issue among competing potential solutions. Here, we can run surveys to find the top option a key group wanted for an issue. Alternatively, we might want to know what use cases we should first support for an in-development product feature.

Foundational Understanding - Product Agnostic: Occasionally, rather than iterate on a given product, we might want to launch a new product or launch an existing product in a new market. To garner success, we need product market fit and a foundational understanding of existing user behavior and top use cases, so that we might build the optimal product experience. This type of research can also help prioritize where we should next launch a product.

Top Predictive Features for a Key User Behavior: Another type of UXR we might want to perform is trying to better understand what characteristics best predict a desirable user behavior or user sentiment. For example, we might want to know what types of product content predict increased user engagement or improved user sentiment.

Deliverables

For quantitative UXR, my typical deliverables include both research decks and short notes or articles highlighting key product takeaways, recommendations, and insights. Depending on the study these might be very tactical recommendations or represent a larger strategic vision for where we should take the product or invest to maximize future success. I also document my code, queries, and other key details for reproducibility, if needed. Understandably, my UX research deliverables are all confidential, proprietary information. However, I have many open source or published papers, decks, and code from academic research, available in my portfolio as well as Github.

Follow @Joshua_Mausolf Follow @jmausolf