01 — Context
HireHealth is a recruitment agency in Coral Gables, Florida, specialising in healthcare and academia. It works with two audiences at once: institutions that need to hire and professionals looking for work.
The brief was for one site to speak just as precisely to people hiring as to people job hunting, without splitting the brand in two or losing trust, and for each audience to find its next step without detours. I designed the UX/UI and built the site in WordPress and Elementor.
02 — My role
I worked on the project through Aquinas Network SAS. I defined the information architecture by audience and the order of the home page, and designed and built the home and contact pages. I set up GA4 and reviewed heatmaps in Clarity. The client supplied the copy and photography.
03 — Diagnosis
Two audiences with opposite intents at the same door
It was a new site, with no previous version to fix: the risk lay in the brief, in ending up with a site that spoke to everyone and no one. My read of the cause: people hiring and people job hunting arrive with opposite questions (can you find me staff? do you have a role for me?), and each audience splits again into healthcare and academia. If the home page doesn’t separate those paths early, every visitor has to read messages meant for someone else before finding their own.
It costs both audiences: whoever runs a clinic or a school, who probably decides within seconds whether the agency understands their sector, and the professional, who is looking for openings, not a sales pitch. And it costs HireHealth, which receives enquiries without knowing which side they come from.
04 — Decisions
A. One button per audience on the first screen
- Problem
- The site speaks to two audiences with opposite intents: people hiring and people looking for work.
- Decision
- The headline bridges talent and opportunity, and the hero offers two buttons, one per audience: the filled "Find the Best Talent" leads to the employer page and the white "Explore Career Opportunities" to the job seeker page.
- Why this one
- I weighed a splash screen to pick a profile ("I’m hiring" or "I’m looking for work"); I chose two buttons in the hero because that screen added a step before the offer. The cost: one audience weighs more than the other, since the filled button is the employers’ one.
- What I expected
- Visitors recognising themselves in one button and taking their path without reading the rest of the home page.
B. A menu that names the audiences, not the services
- Problem
- A menu by service (staffing, academic recruitment) makes visitors translate the offer before finding their own path.
- Decision
- Five items: Home, About, Job Seeker, Employer and Contact Us, with the social links apart. On mobile, the same five in a drop-down panel; in both, the header stays fixed on scroll.
- Why this one
- I weighed a menu by service; I chose to name who each page is for because visitors know who they are before they know what they need. The cost: the menu doesn’t say what each audience is offered until they click.
- What I expected
- Each audience getting back to its path in one click from any page.
C. Two parallel cards, with the sector inside each audience
- Problem
- Each audience splits again into healthcare and academia; four paths on the home page dilute the choice.
- Decision
- A block with two equally weighted cards, Employers and Job Seekers. Each separates healthcare and academia into two paragraphs and ends in a single link: "Hire your next team member with us" or "Search for your next job".
- Why this one
- I weighed four cards, one per audience and sector, and also tabs per audience; I chose two parallel cards because the first multiplied the choice and the second hid one audience behind a click. The cost: long copy in each card, which stacks into two tall blocks on mobile.
- What I expected
- The audience being the decision and the sector a detail, with one link per audience.
D. The same fork near the end of the home page, in reverse order
- Problem
- Anyone who scrolls to the end without choosing has to go back up to find their path.
- Decision
- Before the blog, a block repeats both paths as photo cards: first "Job seekers", with "Explore job opportunities", then "Employers", with "Discover staffing solutions".
- Why this one
- I weighed closing with a generic contact block; I chose to repeat the choice in reverse order because that way job seekers, second in the hero, came first. The cost: the redundancy of presenting the same choice twice on the home page.
- What I expected
- Anyone who finished reading having both paths at hand without going back to the hero.
E. One form, with the audience as a field
- Problem
- Both audiences end up at contact, and the team needs to know which side each enquiry comes from.
- Decision
- The contact page asks for name, phone, email, ZIP code and job title, plus a "Solution Of Interest" selector with two options, Employer or Candidate. Alongside sit the company’s phone, email and office. All three buttons on the employer page lead here.
- Why this one
- I weighed one form per audience; I chose a single one, with the audience as a field, because that was enough to tag each enquiry without duplicating pages. The cost: both audiences answer the same questions, "Job Title" means something different to each, and the audience is chosen last.
- What I expected
- Enquiries arriving tagged by audience, with no call needed to find out.
05 — System
The result was a home page organised by audience: a hero with one button for each (Fig. 0), a menu that names them and stays fixed on desktop and mobile (Fig. 1), a block of parallel cards (Fig. 2), a second fork before the blog (Fig. 3) and a shared contact form, with the audience as a field (Fig. 4).
The scope ended at delivery and instrumentation, not impact reporting, so this case reports no figures.
06 — What I’d test now
Today both audiences end up at the same form, and the audience is chosen in the last field, under a product label: "Solution Of Interest" (Fig. 4). I’d test having visitors who arrive from the employer page find that choice already made.
- Hypothesis
- If the buttons on the employer page open the form with "Employer" preselected, and the selector asks in plain language whether the person wants to hire or is looking for work, submissions from visitors arriving from that page rise and fewer enquiries arrive with the wrong audience.
- Primary metric
- Form submissions per session that passes through the employer page, measured in GA4.
- Secondary
- Share of enquiries with the wrong audience, according to the team’s review.
- Control
- The current form, with the empty selector last. 50/50 split until the sample size calculated from the GA4 baseline is reached, not until a date.