Defining tasks customers complete for a JTBD
In this generative UX research project, I led an initiative to uncover the specific tasks customers complete when trying to achieve one Job-to-be-done identified by Amazon Business. Partnering with the CXO (Customer Experience Outcomes) team, I designed and executed qualitative research with business procurement professionals, synthesized findings into a validated task framework, and delivered strategic recommendations that now inform how teams can benchmark experience quality and evaluate future designs.
SCOPE
Generative UX research, qualitative interviews, task analysis, research synthesis, framework development
TOOLS
Dovetail, Figma, UserTesting, Qualtrics, Excel
TIMELINE
12 weeks
Background
The CXO team had identified that one undisclosed JTBD consistently scored Not Good Enough (NGE) across customer segments. However, there was no shared understanding of what users actually needed to do to achieve that outcome.
Partering with internal stakeholders helped to clarify the core research problem: without a defined task framework, teams could not meaningfully measure usability, benchmark current experiences, or evaluate whether future designs were improving customer understanding.
I structured this work into three phases:
Generative discovery through interviews
Task synthesis and prioritization
Strategic recommendations for integrating tasks into future research workflows
Goals
Identify the key tasks customers complete when completing the JTBD
What actions do users naturally take during this JTBD?
Determine which tasks are most critical to successful decision-making
Which steps most strongly influence confidence and choice?
Clarify which tasks the organization can meaningfully support through design
Where can product and experience improvements drive the most impact?
Format
Phase 1
Developed a stakeholder alignment workshop to educate on the need for tasks and collect thoughts, expectations, and prior JTBD knowledge
Conducted generative, in-depth qualitative interviews (IDIs) with procurement decision-makers
Phase 2
Led task elicitation and synthesized a composite task list across participants
Compared participant-generated tasks against stakeholder-hypothesized tasks
Phase 3
Performed thematic synthesis to capture broader attitudes toward trust, risk, and decision-making
Developed strategic recommendations for operationalizing this framework in evaluative research
Rationale
Because the organization lacked foundational understanding of how customers accomplish this outcome, I intentionally chose generative qualitative methods. Interviews allowed participants to describe real-world behavior in their own language rather than reacting to predefined assumptions, reducing internal bias and increasing validity.
Audience
Interviews: n = 12
8 existing customers
4 non-customers
U.S.-based business procurement decision-makers
Small Market Segment (6–99 employees)
Insights
Discovery Phase
This JTBD is an 18-task process, not a single decision
Through synthesis, I identified a composite list of 18 tasks participants complete when completing the JTBD. This reframed the customer goals as a multi-step behavioral journey rather than a single interaction.
Trust is the foundation of decision-making
Across interviews, I consistently observed that participants’ behavior was shaped by trust: concerns about scams, reliability, and supplier stability strongly influenced their JTBD behavior.
Internal assumptions did not fully align with user reality
By comparing stakeholder-generated task lists with participant-generated tasks, I identified key mismatches (specific examples not publicly available).
Stakeholder Workshops
After completing synthesis, I translated findings into stakeholder-ready artifacts and facilitated alignment conversations with CXO partners and internal teams. Together, we aligned on:
The validated list of 18 customer tasks
Which tasks were most critical to user success
Where experience improvements could meaningfully influence outcomes
This resulted in a shared, user-grounded framework rather than competing assumptions across teams.
Research Validation
Rather than validating a UI prototype, I focused on validating a research framework that could scale.
I proposed and documented how this task framework could support:
Usability benchmarking of current experiences
Task-based evaluation of new designs
RITE (rapid iterative testing and evaluation) approaches
Longitudinal tracking of CXO improvement
I also outlined a practical scoring model (task success + perceived quality) to connect future research directly to measurable experience outcomes.
This positioned the work not just as a one-off study, but as infrastructure for stronger research practice over time.
Conclusions
This project transformed an abstract JTBD metric into a concrete, user-validated framework grounded in real behavior. By defining and validating customer-generated tasks, I helped create a durable foundation for stronger evaluative research, clearer design prioritization, and tighter alignment between user needs and organizational goals.
More importantly, this work demonstrates how I operate as a researcher: reducing ambiguity, challenging assumptions, structuring complexity, and designing research systems that scale beyond a single study.
Limitations
Sample limited to Small Market Segment participants
U.S.-only participants
Qualitative study not designed to measure task prevalence
Could not validate at scale with quantitative and evaluative research
