UX Research · B2B SaaS · Mixed Methods · Behavioral Analytics · Research Operations

Surfacing friction across a four-persona platform through sequenced research.

Surfacing friction across a four-persona platform through sequenced research.

Client

Staffing Referrals

Role

UX Researcher

DELIVERABLES

Research synthesis & recommendations

Findings

Confidential

Research Approach — Sequential Phases

1

2

3

4

5

Behavioral Analysis

HotJar session recordings

All 4 personas

Survey + Interviews

Self-report + market reviews

Recruiters

Competitive Benchmarking

10 companies B2C + B2B

Ambassador behavior

Interviews Round 2

Workflow-focused questioning

Recruiters

Communications Audit

Highest-priority messages complete

All 4 personas

1

2

3

Behavioral Analysis

HotJar session recordings

All 4 personas

Survey + Interviews

Self-report + market reviews

Recruiters

Competitive Benchmarking

10 companies B2C + B2B

Ambassador behavior

4

5

Interviews Round 2

Workflow-focused questioning

Recruiters

Communications Audit

Highest-priority messages complete

All 4 personas

1

2

3

4

5

Behavioral Analysis

HotJar session recordings

All 4 personas

Survey + Interviews

Self-report + market reviews

Recruiters

Competitive Benchmarking

10 companies B2C + B2B

Ambassador behavior

Interviews Round 2

Workflow-focused questioning

Recruiters

Communications Audit

Highest-priority messages complete

All 4 personas

The Challenge

Staffing Referrals is the #1 referral platform for staffing agencies. The platform serves four personas, admins, recruiters, ambassadors, and candidates, each with different workflows, motivations, and relationships to the product.

The core research priority was recruiter adoption. Early behavioral analysis flagged ambassador activity as a factor in recruiter experience, so the scope expanded to keep both in view rather than treat recruiter friction in isolation.

Designing the research itself was part of the work. Four personas with different access patterns and different willingness to engage meant no single method would reach all of them, and the people most affected by adoption friction were the least likely to respond to outreach. The research used a sequence of methods, with each phase building on what the previous one had shown.

Each method was chosen to reveal what a given persona could uniquely show or say.

Strategy

Behavioral analysis came first to map how each persona interacted with the product and establish a baseline before interview questions were written. It also reached all four personas without depending on participation, which mattered given the recruiter response patterns the later phases would confirm.

Methods by persona

Admins

Behavioral analysis

Primary workflow data from session recordings.

Recruiters

Survey + interviews

Self-report data paired with market reviews. Low response rates themselves signaled the adoption problem.

Interviews round 2

Workflow-focused questioning to go deeper on day-to-day work.

Ambassadors

Behavioral analysis

Most visible in session recordings.

Competitive benchmarking

Referral motivation patterns across ten B2C and B2B companies.

Candidates

Behavioral analysis

Mobile behavior visible in session recordings.

Phase 1 — Behavioral Analysis (HotJar)

Across all personas

I started by mapping each persona's likely tasks, then used URL keywords tied to those tasks to filter HotJar recordings by persona. About 50 recordings were reviewed across the four groups.

The communications audit reached all four personas. It was the only phase scoped to do so.

Ambassador and candidate activity was the most visible in the data, consistently on mobile. Recruiter activity showed up only on a narrow slice of pages, consistent with the adoption concern that prompted the research.

Phase 1 — Behavioral Analysis (HotJar)

I started by mapping each persona's likely tasks, then used URL keywords tied to those tasks to filter HotJar recordings by persona. About 50 recordings were reviewed across the four groups.

Ambassador and candidate activity was the most visible in the data, consistently on mobile. Recruiter activity showed up only on a narrow slice of pages, consistent with the adoption concern that prompted the research.

Phase 2 — Recruiter Survey, Market Reviews, and Interviews

The thin recruiter footprint in HotJar was what prompted Phase 2. Behavioral data alone wasn't going to be enough to understand recruiter experience, so the work expanded to three sources of self-report.

A survey went to active platform users to capture articulated friction across the recruiter base. Market reviews pulled recruiter feedback from external review sites and forums as additional input. In-depth interviews followed with recruiters willing to engage, going deeper into the patterns from the survey and review data.

Recruiter response rates to direct outreach were low across two rounds. The three sources together gave the phase reach the interviews alone couldn't have provided.

Phase 3 — Competitive Benchmarking

With ambassador activity flagged as a factor in recruiter experience, competitive benchmarking turned to referral programs running outside the staffing category. Ten companies were analyzed, a mix of B2C and B2B. The analysis looked at how each handled referral CTA placement, reward structure, sharing methods, and dashboards.

The output was a swipe file of patterns and opportunities, which fed into the broader set of recommendations alongside the survey, market reviews, and interviews. With ambassador research in hand, the next phase returned to recruiters.

10 companies. B2C and B2B. Because referral motivation is a human pattern, not an industry-specific one.

Phase 4 — Recruiter Interviews, Round 2

The first round of recruiter interviews directed the following round to focus more deeply on workflows. Four interviews were conducted, exploring what recruiters prioritized in their day-to-day work. The interviews, alongside the other phases, pointed to communications as an area worth examining. A notifications audit became the next phase.

PHASE 5 — COMMUNICATIONS AUDIT

The notifications audit covered roughly 40 communications across the four audiences. Each was reviewed and streamlined to be more efficient and actionable, stripping unnecessary or competing elements so the user's next single action was clear. Related communications were evaluated together for consistency across the set, and the redesigned communications were matched to the platform's new branding.

This phase prioritized the messages users encountered most often and those carrying the highest risk of introducing friction. The deliverable was a redesigned set of communications, accompanied by a spreadsheet of decisions and rationale.

Outcome

The engagement covered five phases of research across four personas, beginning with behavioral analysis of platform usage and ending with a notifications audit and redesign. Communications emerged as a point of friction across research and leadership's standing concerns, which directed Phase 5 to address it across the full audience set.

Findings are confidential at the client's request. Research design and method selection are available in full.

The deliverables are currently under leadership review.

Next Project

View Case Study

A concept feature for an EdTech platform, redirected by what families experienced.

UX Research • EdTech • Mixed Methods • Feature Addition • Concept Design

Next Project

View Case Study

A concept feature for an EdTech platform, redirected by what families experienced.

UX Research • EdTech • Mixed Methods • Feature Addition • Concept Design