EMPLOYEE EXPERIENCE · IN-HOUSE UX RESEARCH
Identified pain points across Enterprise Sales Representatives and mapped opportunities to AI use cases for the largest global medtech company.
ROLE
UX Project Manager
TIMELINE
8 weeks
METHODS
Stakeholder Interviews · Pain Point Mapping
01 / PROBLEM
Enterprise Accounts sales representatives at a major medical device company manage large, multi-stakeholder accounts across a broad product portfolio. Fragmented systems and manual processes meant a substantial share of their week went to administrative work rather than to customers. This was a major problem since revenue is won through the time sales reps. have with the customers.
The company already had an internal catalog of 89 AI use cases, ranked and scored. So the problem wasn't a shortage of ideas.
The problem was that we needed to identify where the sales representatives were spending the most time away from making deals, and giving them solutions they would actually care about.
Nobody had checked whether the top-ranked opportunities identified internally had matched the lived reality of the Enterprise Accounts team specifically, or whether the friction those reps. actually felt was even represented in the list.
02 / APPROACH
I designed the engagement to be a semi-structured interview format, where we looked at existing past survey data instead of creating our own.
Running more surveys would produce more of the same kind of evidence, so we decided to focus on the conversational interviews. What was missing was the qualitative layer that could tell us whether those rankings held up against how the work actually happens.
So, I approached this study as an opportunity to dig deep into the sales journey of the Enterprise Accounts team, identifying where they were experiencing the most friction, what activities were taking up the most time, and whether those processes should be human-led or automated.
I structured interviews around the four phases of the deal cycle: account review, account planning, contract execution, post-deal growth. This made it easier to locate friction in the process rather than collect data as a general list of complaints.
03 / PROCESS
01 - Scope
Defined the engagement around one persona (the Enterprise Accounts team) rather than the full sales population, so findings could be specific enough to act on.
02 - Interview
Led 15 stakeholder interviews across director, distributor, and sales roles on the Enterprise Accounts team, probing deal timelines, top workflow frustrations, and current AI tool use.
03 - Map friction to phase
Synthesized findings into a pain point matrix plotting each issue by frequency and severity, then organized the top pain points under each of the four deal phases.
04 - Quantify the qualitative
Tagged every pain point with how many of the 15 interviews independently surfaced it, producing a validation count that made qualitative findings legible to a quantitative audience.
05 - Validate the catalog
Tested the top 25 ranked use cases against interview evidence: 13 directly validated, 7 partially, 4 with no field alignment at all.
06 - Rescore
Rebuilt the priority ranking using criteria weighted for Enterprise Accounts. Use cases that ranked low in composite scoring rose sharply after qualitative validation.
06 - Recommend
Narrowed to 12 validated use cases that were categorized as either immediately deployable or most strategic. Mapping recommendations along a 24-month implementation timeline and a risk/mitigation plan.
04 / ARTIFACTS
See my finished report below!
05 / IMPACT
We recommended twelve validated use cases, projected at roughly 26 hours per month returned to each rep. This was time currently lost to manual aggregation, contract handoffs, rebate tracking, and quarter-end preparations.
But the more useful output was the re-scoring. Several of the highest-value opportunities for this team were sitting in the middle of a list that had been ranked correctly for the average seller and therefore incorrectly for this one. The biggest takeaway was that a composite ranking was not enough to identify value in where to integrate AI, and that actual user conversations and qualitative data were needed.
After delivery, the three areas our interviews identified as the deepest friction matched first-pass findings from the client's own consolidation surveys: rebate management, manual data and reporting, and contracting.
We also flagged what the recommendations couldn't fix on their own: existing AI tools at the company were already available and already underused. Thus, understanding AI literary levels, resistance drivers, and change readiness across the organization can improve success and future company-wide AI implementations.
06 / LEARNINGS
1) Validation is a real research contribution, not a lesser one. There is real value in coming to the same conclusion as consultants from ZS and Deloitte from our own data. This provided our team with the confidence that our hard work became another stepping stone towards company-wide AI adoption.
2) As a researcher, it's important to stay human-centered. While learning more about the sales workflow, I found that it's important to find the balance between where AI can be implemented and where it should be implemented. Understanding the value a sales rep brings to the selling process helped me identify what AI use cases should be left to the humans.
3) The hardest problems in AI adoption aren't technical. Every rep we spoke to had access to AI tools. The real challenge was understanding how to use it in a smart and efficient way. Digging deeper into the mental models of these sales reps allowed us to identify what the real issue was.