Usually because the page tells them nothing they could not guess. A list of open roles, a stock photo of people laughing near a laptop, and a sentence about being passionate does not help anyone decide. Good candidates have options, so they filter fast, and a vague page gets filtered out first.
The teams we work with are often surprised by this. They put real effort into the homepage and the services pages, then treat careers as an internal admin page that HR owns. It gets built once, linked from the footer, and never looked at again.
That is a mistake, because a careers page is a conversion page. The visitor has a decision to make and you are trying to help them make it. Everything you know about designing a landing page applies here.
It should help the right people say yes and the wrong people say no. Both halves matter. A page that attracts a hundred unsuitable applications has cost your team more time than it saved. Being specific about who this suits is a feature, not a risk.
We think of it as three jobs in order. Tell someone what the company actually does in plain language. Show them what working there is like, honestly. Then make applying to a specific role feel low friction and worth the effort.
Most careers pages only attempt the third job, and they attempt it with a list of links. The first two are where the persuasion happens, and skipping them means every candidate arrives at your job listings with no context for whether they want the job.
A clear statement of what the company does, who you are hiring, and where the work happens. Remote, hybrid, or office, and which country or time zone. Those two facts filter more candidates correctly than anything else on the page, so hiding them lower down wastes everyone's time.
Location and work model are the most common reason a good candidate stops reading. If someone has to open three job listings to discover the role is office based in a city they do not live in, you have annoyed a person who might have been right for a future role.
Keep the headline honest rather than clever. "We build payment tools for small retailers, and we are hiring three engineers in Melbourne" outperforms "Join the journey" every single time, because it answers the question the visitor came with.
The rest of the trust building can borrow from the page you already have. A careers page and an about page do overlapping work, and the principles carry across, which we covered in our post on what makes a good about page.
Lead with what the person will do in their first six months, not with a paragraph about your company. Candidates read the responsibilities first regardless of what order you put things in, so putting the company boilerplate at the top just adds a scroll before the useful part.
Cut the requirements list down to what is genuinely required. Long lists of desirable skills discourage exactly the careful, self-assessing candidates you want, while the overconfident apply anyway. If a skill is teachable in the first month, it is not a requirement.
Be concrete about the day to day. Which team, how many people, what tools, what the first project is likely to be, and who they report to. Every specific detail is something a candidate can picture themselves doing, and vague language gives them nothing to hold onto.
If your listings are long, collapse the secondary sections rather than deleting them. Nielsen Norman Group's guidance on progressive disclosure is to "disclose everything that users frequently need up front" and move the rest to a secondary layer. Responsibilities and salary belong up front. Benefits detail and legal boilerplate can sit behind a toggle.
Yes, and we will take the strong position. Publishing a range costs you a little negotiating leverage and saves both sides an enormous amount of wasted time. A candidate who would never accept your range finds out in ten seconds instead of after three interviews.
The argument against is usually that competitors will see it. They will, and they already have a good estimate. The argument that internal staff will compare is more serious, but that is a compensation problem you already have, and hiding the number does not solve it.
What we notice in practice is that pages without a range get treated as a warning sign. Candidates assume the number is bad, because a good number would be advertised. Silence is not neutral, it is read as a signal, and usually not a flattering one.
If you truly cannot publish a range, publish something. A band, a starting point, or an honest note about how compensation is set is better than nothing at all.
Show the work and the people, not the office furniture. A short description of how a project actually runs, who decides what, and how feedback happens tells a candidate far more than a gallery of the kitchen area. Specific beats aspirational every time.
Real photographs of the actual team do a lot of work here. Stock photography on a careers page is uniquely damaging, because the reader is trying to judge whether these are people they want to spend their days with, and you have shown them strangers. Any honest photo beats a polished fake one.
The most persuasive thing you can publish is something slightly unflattering. Say what is hard about the job, what the company is not good at yet, or what kind of person struggles there. Candidates read that as confidence, and it filters out people who would have left in four months anyway.
Employee words help if they are real. A two-sentence quote from a named person about a specific project beats a paragraph of polished testimonial that could have come from anywhere.
If you post individual jobs, yes. Google supports JobPosting structured data, and it requires five properties: datePosted, description, hiringOrganization, jobLocation, and title. Adding them makes your listings eligible for Google's job search experience rather than sitting invisible behind your careers page.
The rule people break is about expired roles. Google's documentation is unusually blunt here, stating that jobs no longer open for applications "must be expired" and that "failure to take timely action on expired jobs may result in a manual action". A manual action is a penalty, not a warning.
Google gives three acceptable ways to expire a listing. Populate the validThrough property with a date in the past, remove the page entirely so it returns a 404 or 410 status, or strip the JobPosting structured data from the page. Leaving a filled role live with valid markup is the one option that is not allowed.
This is why we build careers listings on a CMS collection rather than as hand-built pages. A collection lets you set a closing date once and have the markup and the page behave correctly without anyone remembering to tidy up. If structured data is new to you, our guide to schema markup and how it boosts your SEO covers the basics.
Short enough that a good candidate finishes it on their phone. Name, contact details, the role, and a way to attach something. Everything else can wait until you have decided you are interested, and most of it can be asked in a conversation.
Baymard Institute's checkout research is the closest reliable data we have on form length, and the pattern transfers. Their measurements put the average checkout at 11.3 form fields in 2024, while their position is that "most sites need only 8 form fields in total for a checkout flow". Their point is that extra fields raise "the perceived and actual complexity" of the task.
A job application is a much bigger ask than a checkout, and candidates are usually applying to several places in one sitting. The application that takes four minutes gets finished. The one that requires re-typing a CV into separate fields gets abandoned, and it gets abandoned disproportionately by people who have other offers.
Applicant tracking systems are often the culprit rather than the design. Greenhouse, Lever, Ashby, and Workable all let you configure which fields are required, and the defaults are usually greedier than they need to be. Someone has to go in and turn things off. We go deeper on this in our guide to designing forms people actually finish.
Tell them what to expect and when. A confirmation screen that says what the next step is and roughly how long it takes turns an anxious wait into a normal one. Silence after an application is the single most common complaint candidates have, and it is entirely a design problem.
Set the expectation you can actually meet. If you review applications weekly, say so. A promise of two days that turns into three weeks does more damage than an honest promise of three weeks, because now the candidate has learned something about how the company keeps its word.
Then close the loop on rejections. Every person you turn down is a potential future applicant, a customer, or somebody a friend asks about you. A short, prompt no is remembered kindly. No reply at all is remembered too, just differently.
Open your careers page on a phone and read it as a stranger. Can you tell what the company does, where the work happens, and what you would be doing, within about thirty seconds? If any of those three is missing, fix that before you touch anything visual.
Then check your job listings for expired roles with live structured data still on them. That is the one item on this page with a genuine penalty attached, and it is usually the fastest thing on the list to fix.
If you are rebuilding a careers section and want it to work as hard as the rest of your site, we are happy to help you think it through. Careers pages are one of the most neglected pages we get asked to look at, and usually one of the easiest to improve. Let's talk, and you can find us at phoenix.studio.
Tell us where you want to go. We'll tell you how we'd get you there.