"They didn't want more options. They wanted someone to tell them: this one. She's the right fit."
Business owners coming to PONS weren't lazy or indecisive; they were scared. Scared of wasting money on the wrong hire. Scared of onboarding someone remotely who wouldn't deliver. When they landed on a platform with 3,000+ profiles and no clear signal of who to trust, that fear didn't go away; it got louder. The symptom was low conversion. The real problem was a lack of trust, and an experience that accidentally made things worse by throwing too many choices at people who were already anxious about choosing.
PONS had supply. PONS had demand. But the bridge between them was broken. Clients were arriving, browsing briefly, and leaving without hiring. The team spent two full days manually assembling shortlists that still felt untrustworthy, with no visible quality signal on any profile.
Choice overload was the #1 drop-off driver. Clients who browsed more than 20 profiles almost never converted. The more they saw, the less they trusted any single option.
Verified meant nothing without context. A basic ID check wasn't reassuring; it just confirmed the person existed. Clients wanted to know: can this person do the job, and has someone I trust already confirmed that?
The team's expert filtering was invisible. Hours of manual shortlisting were happening behind the scenes, but clients had no idea; they assumed they were looking at a raw, unfiltered list.
SDR roles carried the highest anxiety. Sales hires are performance-sensitive. A bad SDR doesn't just cost money; they damage pipeline momentum. This segment needed the most reassurance before committing.
The existing tagging system used broad labels like "Sales Representative." These told clients almost nothing about fit. In collaboration with the PM and engineers, I redesigned profile architecture around specific competencies and years of hands-on experience:
For clients: fast scannable signal of depth. For engineering: structured data that made precise automated shortlisting possible for the first time. A progressive disclosure pattern kept rep completion rates high without adding friction.
A two-tier trust system that made quality visible at every touchpoint, shaped by PM conversations and engineering input on what verification workflows were realistically buildable.
The design principle was earned visibility. Vetted profiles surfaced higher in shortlists, carried a distinct visual treatment, and were the only ones eligible for the premium agent workflow. For clients, the Vetted badge acted as a proxy for a trusted referral.
For clients who needed the highest support, including first-time remote hirers or founders stretched too thin, instead of browsing profiles, the client submits a structured brief: role type, required tools, team pace, culture notes. A PONS agent then returns exactly five Vetted candidates, each with a personal match rationale.
Engineers initially pushed back on scope. Working with the PM, we prioritized a lightweight first version: a simple brief form feeding into the existing agent workflow, with a richer dashboard scoped to a later sprint. It shipped faster and still solved the core problem completely.
"Sometimes the best design is removing the interface entirely. The most anxious users didn't need a better search; they needed someone to take the wheel."
The Vetted badge worked; clients trusted it, conversion improved. But success created its own problem. Every Vetted badge required a live interview and manual data validation. Within weeks of launch, demand outpaced capacity: a six-week backlog of pending interviews. The platform was caught in a paradox: the thing making us credible was also making us look empty.
Keep unvetted reps hidden → talent pool looks sparse → clients assume the platform isn't worth their time and leave.
Let everyone in without checks → clients meet underqualified reps → the boutique reputation unravels quietly but fast.
I designed an Admin-Only "Completed Profile" filter, invisible to clients but critical for the recruitment team. Every rep was required to submit a video introduction before their profile became visible. The team reviewed these internally to fast-track the strongest candidates into the interview queue, a hidden quality layer that kept the public experience clean.
While the backlog was being cleared, the Dedicated Agent service was offered for free during this transition period, absorbing uncertainty and ensuring every client experience felt intentional.
By making Verified-but-not-yet-Vetted reps visible, we accepted a real risk: some clients would connect with reps who didn't yet meet our full boutique standard. That was a conscious call, not an oversight. The alternative, a platform that looked empty, carried a higher long-term risk. Owning that tradeoff clearly also made internal alignment significantly smoother.
Clients who experienced the hands-on agent service during the transition period became some of the most loyal returning buyers. The free period proved the value of the premium service to exactly the clients who needed convincing most.
Trust is a design material. The biggest UX problem wasn't navigation or search; it was anxiety. Designing for trust meant making quality legible, expert judgment visible, and sometimes removing choice altogether.
Collaboration shapes better constraints. Some of the strongest design decisions came directly from PM and engineering conversations. Cross-functional alignment wasn't overhead; it was the work.
The best solutions sometimes live outside the screen. The dedicated agent workflow was a service design decision as much as a product one. Some users needed a human layer, not a better UI.
Own your tradeoffs. Design maturity means knowing what you're giving up, not just what you're gaining. Making the two-tier decision explicitly built more internal trust than any technically perfect solution would have.
Structure unlocks scale. The Skill-Year Matrix compounded across filtering, shortlisting, client confidence, and team efficiency, all at once.