A Q&A field guide to site architecture, navigation, the split-domain question, first-screen design, and the quiet compromises that drain both pipelines

Picture a restaurant where half the room came for dinner and the other half came to buy the restaurant. That is a staffing website. The candidate wants a job near home, this week, from a phone with a cracked screen. The hiring manager wants proof you won’t waste her quarter. In iHire’s 2025 State of Online Recruiting report, 59.7% of employers said job boards send them too many unqualified candidates. Meanwhile Indeed found that 42% of job seekers call lengthy applications the most frustrating part of applying. Employers want a filter. Candidates want a shortcut. Every design decision on your site lands somewhere on that rope.

Q1. Why does the website matter more now than it did three years ago?

• The market got smaller and pickier. Staffing Industry Analysts (SIA) projects U.S. staffing revenue of $183.1 billion in 2026 (up 2.4%) and $187.0 billion in 2027, well below the $243.9 billion peak of 2022 (Fig. 1). The American Staffing Association (ASA) reports 2025 temp and contract employment of 9.5 million and sales of $113.5 billion, both down 8.5%. Turnover hit 376%, so you re-recruit your workforce almost four times a year.
• Candidates have options that aren’t you. Bullhorn’s GRID 2025 Talent Trends survey of roughly 2,800 workers found only 35% turn to a recruitment agency first, down from the prior year.
• Buyers decide before they call. Gartner found B2B buyers spend only 17% of buying time meeting suppliers, and just 5% to 6% with any one rep when comparing vendors (Fig. 5). 6sense’s 2024 Buyer Experience Report says 81% already have a preferred vendor at first contact. Your website makes the first pitch, and often the last.

Q2. Who is actually visiting, and how differently do they behave?

• Candidates are high volume, mobile, and impatient. Appcast says about two-thirds of applies come from phones. Indeed says 78% of seekers would apply by phone if it were friendlier. Google found 53% of mobile visits are abandoned when a page takes over 3 seconds to load.
• Employers are low volume and high value. They browse from a desktop at work and rarely buy alone; Gartner counts six to 10 stakeholders per buying group. If you let pageviews vote, candidates win every election and the sales team starves. Segment analytics by path before traffic decides your homepage.

Q3. What does each audience need on the first screen?

Nielsen Norman Group’s 2018 eye-tracking work found 57% of viewing time lands above the fold and 74% within the first two screens (Fig. 4). The Stanford Web Credibility Project (Fogg et al., 2,684 participants) found 46.1% of people’s credibility comments referred to a site’s visual design and look. The first screen has two jobs: route each visitor in one tap, and look worth trusting.

Table 1. The first screen, as each audience sees it

	Employer (the buyer)	Candidate (the talent)
Question in their head	Can these people fill my role faster and better than my team can?	Is there a job for me, near me, and what does it pay?
Must see instantly	Specialties by industry and role, hard proof (fill rate, time to shortlist, client ratings), one clear “Request talent” button	A job search box (keyword plus ZIP code), pay ranges, and a promise like “Apply in 2 minutes”
Usual device and time	Desktop, office hours, often reading with colleagues	Phone, evenings and weekends. PeopleReady saw 55% of app assignments happen outside business hours
Instant turnoff	“People are our passion” copy with no specifics	A login wall before any jobs appear

Q4. How should navigation split them without making either feel like a second-class guest?

• Split at the top, one tap deep. Adecco’s persistent “For Job Seekers” and “For Employers” tabs swap headlines, tools, and even the chat assistant’s role, per HubSpot’s review. Remember the choice in a cookie and keep a visible switch.
• Label by intent. “Find a job” and “Hire talent” beat “Candidates” and “Clients.” Few people describe themselves as a candidate.
• Watch out for the third audience. “Careers” on a staffing site usually means jobs at the firm itself. Move it to the footer as “Work at [Firm]” so job seekers don’t wander into your recruiter openings.
• Route contact by audience. ClearlyRated’s client benchmark shows only 69% of staffing clients get calls and emails returned within 24 hours. Separate forms help.

Q5. Should you split into separate domains?

Usually no. Split the experience and keep the domain, because search authority accrues to one domain. SEO veteran Rand Fishkin compiled 14 case studies where moving content from a subdomain to a subfolder increased search traffic. B2B agency Obility reported a 75% organic lift after moving a client’s blog from a subdomain onto the main domain (Fig. 7).

Table 2. Four ways to structure the site

Structure	Works well when	Watch out for	Example
One domain, audience tabs	Most firms, one brand, one sales team	Homepage that tries to talk to both at once	Adecco (persistent tabs)
One domain, subfolders (/jobs, /hire)	You want every job page to build domain authority	Keeping ATS jobs on your domain takes setup	The SEO default
Subdomain (jobs.firm.com)	Your ATS vendor hosts the job board	Authority splits; jarring handoff	Common in mid-size firms
Separate apps or brands	The candidate side is a product people use daily	Two budgets, two roadmaps, two brands to maintain	PeopleReady JobStack

PeopleReady is the exception: its JobStack platform runs separate worker and business apps. By 2019 it filled a job every 9 seconds, handled 43% of PeopleReady shifts, and served nearly 17,000 companies and 68,000 daily active workers (TrueBlue). If your candidates open your product every day, split. If they visit twice a year, don’t.

Q6. What does the candidate path need in order to convert?

• Short applications. Appcast data shows applications finished in under 5 minutes convert at 12.47%, against 3.61% for those taking over 15 minutes (Fig. 2). Past 5 minutes, drop-off runs 50% to 75%. Skip the account creation until after they apply.
• Fewer screener questions. Indeed reports 30 screener questions cut applications by 50%, and more than 45 cut them by almost 90% (Fig. 3).
• Pay up front. CareerPlug’s 2025 research found 47% of seekers want salary information before applying. Appcast’s 2026 data says job titles of 4 to 6 words perform best.
• Structured job data. Google’s case studies show ZipRecruiter got 4.5x higher organic conversion and 35% more non-branded traffic after adding JobPosting markup. Monster India saw job-page organic traffic rise 94% and applications rise 10%.

Q7. What does the employer path need in order to convert?

• Self-serve proof. Gartner’s 2025 survey of 632 buyers found 61% prefer a rep-free buying experience and 73% actively avoid suppliers who send irrelevant outreach. Build specialty pages with roles filled, time to shortlist, and case studies with real numbers.
• Publish your ratings. ClearlyRated’s industry Net Promoter Score benchmark sits at 36% from clients and 30% from placed talent, while Best of Staffing winners averaged 77% and 73% in 2023 (Fig. 6). If you beat the average, say so on the first screen of the employer path.
• Use assets that serve both sides. Salary guides are rare shared content. Hiring managers use them to budget, candidates use them to negotiate. Michael Page and Robert Half feature theirs prominently. Then add a short “Request talent” form: role, location, headcount, start date.

Q8. Which compromises quietly cost you both audiences?

The compromise	Why it looks smart	What employers lose	What candidates lose
Rotating hero slider (“Find a job” then “Hire talent”)	Feels fair to both	Lands on the job slide and leaves	Lands on the sales slide and leaves
“We connect great people with great companies”	Inclusive, safe	No sign of specialty	No sign of jobs
One contact form for everyone	Simple to build	Leads buried under resume follow-ups	Messages land with sales and sit
Job board on the ATS vendor’s domain	Cheap and quick	Search authority goes elsewhere	Jarring handoff, second login
Account required to apply	Cleaner database	Fewer people to submit	Up to 75% drop-off past 5 minutes
Stock photos of handshakes	Looks corporate	Signals “generic vendor”	Doesn’t look like their workplace

Q9. How do you know whether the site is working for both sides?

• Run two funnels. Candidates: job view to completed apply, benchmarked against Appcast’s 6.1% apply rate for 2024. Employers: specialty page view to request to closed deal.
• Track what happens after the click. Bullhorn found 72% of candidates quit a recruiter over slow or poor communication (Fig. 8). CareerPlug says 26% declined offers over poor communication. Then apply to your own job on a phone with a stopwatch running. If you get bored, so did they.

Q10. What is the whole blueprint in one breath?

1. Split the experience at the top, keep one domain, and let each audience reach its path in one tap.
2. Give candidates search, pay, and a sub-5-minute mobile apply. Give employers specialties, proof, and a short request form. Host jobs on your domain with JobPosting markup.
3. Publish ratings and numbers, since buyers spend only 17% of their time with vendors. Then kill the slider, the shared form, and the login wall. One site can work like two.

Sources

1. Appcast, 2025 Recruitment Marketing Benchmark Report: https://www.appcast.io/the-2025-recruitment-marketing-benchmark-report-is-live/
2. Appcast, job ad and apply-time research (12.5% vs. 3.6%): https://www.appcast.io/how-to-write-a-job-ad-that-converts/
3. Appcast, drop-off by apply time and question count: https://www.appcast.io/create-a-remarkable-candidate-experience/
4. Appcast, mobile vs. desktop trends: https://www.appcast.io/mobile-versus-desktop-trends-in-job-search/
5. Appcast, job ad content best practices (2026): https://www.appcast.io/job-ad-content-best-practices/
6. Indeed, why great applicants abandon the process: https://www.indeed.com/lead/reasons-great-applicants-abandon-the-process
7. iHire, The State of Online Recruiting 2025: https://www.ihire.com/resourcecenter/employer/pages/the-state-of-online-recruiting-2025
8. Staffing Industry Analysts, U.S. Staffing Industry Forecast, Sept. 2026: https://www.staffingindustry.com/research/research-reports/americas/us-staffing-industry-forecast-september-2026-update
9. StaffingPulse, 2022 peak and pre-pandemic market size: https://www.thestaffingpulse.com/news/revenue-decline-2025.html
10. American Staffing Association, Q4 2025 employment and sales: https://americanstaffing.net/posts/2026/03/30/employment-and-sales-rebound-in-q4/
11. Bullhorn, GRID 2025 Talent Trends Report: https://www.bullhorn.com/grid/2025-talent-trends/
12. Bullhorn press release (72% communication finding): https://www.businesswire.com/news/home/20250924249047/en/The-Human-AI-Partnership-New-Data-From-Bullhorn-Reveals-How-AI-Can-Save-the-Candidate-Experience
13. Gartner, B2B buyers prefer rep-free buying (2025): https://www.gartner.com/en/newsroom/press-releases/2025-06-25-gartner-sales-survey-finds-61-percent-of-b2b-buyers-prefer-a-rep-free-buying-experience
14. Gartner buying-time breakdown (17% with suppliers): https://www.slideshare.net/slideshow/gartner-buyer-enablement-winmoreb2bsalesdealspdf/258207352
15. 6sense, 2024 Buyer Experience Report: https://www.businesswire.com/news/home/20241009142556/en
16. Nielsen Norman Group, Scrolling and Attention: https://www.nngroup.com/articles/scrolling-and-attention/
17. Stanford Web Credibility Project: https://credibility.stanford.edu/
18. Google mobile speed research via Marketing Dive: https://www.marketingdive.com/news/google-53-of-mobile-users-abandon-sites-that-take-over-3-seconds-to-load/426070/
19. ClearlyRated, NPS benchmarks for staffing: https://www.clearlyrated.com/blog/nps-101-staffing-recruiting
20. ClearlyRated, client satisfaction driver benchmarks: https://knowledge.clearlyrated.com/client-services/product/staffing-industry-benchmarks/
21. Medix, Best of Staffing winner averages: https://www.medixteam.com/blog/understanding-nps-and-its-critical-role-in-staffing-partnerships/
22. TrueBlue, PeopleReady JobStack results: https://investor.trueblue.com/news-events/press-releases/detail/314/peoplereadys-jobstack-app-fills-a-job-every-9-seconds
23. Google Search Central, ZipRecruiter case study: https://developers.google.com/search/case-studies/ziprecruiter-case-study
24. Google Search Central, Monster India case study: https://developers.google.com/search/case-studies/monster-india-case-study
25. Obility, subdomain to main domain migration: https://www.obilityb2b.com/blog/our-work/case-studies/blog-subdomain-to-main-domain-migration-how-obility-did-it-and-generated-a-75-organic-traffic-increase/
26. Rand Fishkin, 14 subdomain vs. subfolder case studies: https://twitter.com/randfish/status/974011592902770688
27. HubSpot, recruitment website design examples (Adecco): https://blog.hubspot.com/website/recruitment-websites-design
28. CareerPlug, Candidate Experience Report 2025: https://www.careerplug.com/candidate-experience-statistics/