By 2027, a lot of roles will have drifted far enough from their postings that a keyword search built in 2019 turns up the wrong people, or nobody at all.
The repair starts further upstream than most teams look: in how roles get named, what the posting actually asks for, and whether anyone has mapped this year’s tasks onto next year’s skills.
Recruiters feel it first, but the cost lands wider.
A candidate trying to decide which skills to bet on reads these postings for signal and gets noise. A team that keeps writing reqs for the job it used to have keeps hiring for the past, then pays again to backfill the gap.
The 2027 Gap, Defined
The tools that used to sit in side projects now run the main work: generative AI, automation, the systems that move and store a company’s data.
Whole industries are rebuilding at once, pushed by the drive to cut emissions, to put supply chains online, and to organize around products rather than departments. The workforce has shifted with it, spread across generations and locations the old role templates were never written for.
The gap is plain enough to state: the work is changing faster than the names, structures, and skills built around it can keep up.
Where Job Descriptions Fail
A job description either anchors a team to the past or points it at the work coming.
Most anchor. They list a tech stack that’s already aging and credentials that proxy badly for capability. They count years of experience instead of evidence someone can do the thing.
Catch-all titles smear accountability across a team, and phrases like “fast-paced environment” or “self-starter” paper over what the role actually involves.
A degree requirement where a skills check would do screens out people who could do the job, and screens them out before anyone reads their work.
There’s a deeper miss underneath all of that. Plenty of JDs have no clear person in mind. Nobody drafting them can picture who the role serves or who sits next to it, so the posting ends up addressed to everyone, which means it lands on no one in particular.
The result reads generic because it was written generically.
Writing Roles Toward the Work
No role is future-proof. A job description can still show where a role is heading, as long as it comes from real research and not from habit.
Start with what the work needs, not the tools in use today. Tools change fast, and a description tied to them ages quickly. One that names the skills the job actually calls for holds up far longer.
Write it with the right people. A manager and someone who does the work can describe a real day: the tasks, the decisions, what takes up the time. That beats a wish list. HR can add which skills line up with the people who do the job well.
Dr. Theerapong Poonyakariyagorn, founder of Interplast Clinic, runs a cosmetic surgery practice that has had to rewrite roles as the technology in the field moved on.
He says, “Even in a specialist clinic, the titles stopped matching the work a few years ago. A coordinator on my team now handles imaging software and patient data that nobody in that seat touched when we opened. We learned to build the role around what the person does each day, because a title carried over from how the clinic ran ten years ago only confuses the people we are trying to hire.”
Some skills never fit in a title but still matter. Keeping a team steady through change is one. Working well with new tools is another. Both belong in the description.
Review it on a schedule. Fast-moving roles need a look every few months, steadier ones less often. Each time, the question is the same: which new tools or ways of working actually stuck. The description then gets updated to match the job as it really is.
Rebuilding a Job Description
The starting move is an audit of what the team can do now, set against what the strategy will need twelve to twenty-four months out. Whatever sits in the gap is the hiring-and-upskilling plan, already written.
From there, separate the capabilities a role genuinely can’t open without from the ones that would be nice to have, and tie each one to a task or an outcome.
Where it’s possible, “5+ years of X” becomes “can produce Y using X”, the second version is checkable, the first is a filter standing in for one.
Language does damage, so it’s worth auditing too. Jargon and coded phrasing turn away qualified applicants who’d otherwise apply, and tools that flag biased or confusing wording tend to improve both the size and the fit of the pipeline.
Keep job families and competencies in one source of truth rather than scattered across managers’ drafts, watch how the candidate pool moves when a requirement or a phrase changes, and adjust from what the data shows.
What’s Working Elsewhere
A handful of companies have already moved on this, and their approaches don’t look alike. What they share is a decision to make the real work visible and name the skills plainly.
IBM’s skills-first hiring
IBM pulled degree requirements from more than half its U.S. openings and built routes in through apprenticeships and earn-and-learn programs.
The talent pool widened, and candidates matched to work more cleanly once the credential stopped doing the sorting.
Unilever’s internal talent marketplace
Unilever runs an AI-powered platform that matches work to skills. Teams post projects, employees pick up experience past whatever their title says, and the job content shifts fast without anyone redrawing the org chart every quarter.
Amazon’s upskilling push
Amazon put more than a billion dollars behind upskilling hundreds of thousands of employees by 2025, pointing the development paths at in-demand roles like data science and mechatronics. Clearer role definitions made the transfer legible — people could see how their current skills carried into the next job.
Language and structure matter
Rewriting job ads in specific, inclusive language shows up in the results. Research on gender-coded and exclusionary phrasing finds that plain, welcoming wording pulls a more diverse and more qualified field, which matters most in exactly the fast-changing fields where the gap is widest.
Training After the Hire
A job description is a door. It gets someone into the building; it doesn’t carry them anywhere once they’re in. That’s what training does: stackable credentials, apprenticeships, mentorship, the learning that happens on the actual job, and it’s what turns a decent hire into someone worth keeping for years.
A JD should carry both horizons at once, then: the skills required on day one, and the skills the role will help someone build over the first year. That second line is often what closes a candidate who’s weighing the move, because it tells them the job goes somewhere.
Starting Points
A posting that describes last year’s job loses this year’s candidate. The fix is specific. The role has to match what the job actually involves. The description has to name the skills that will matter next year, not only the ones that matter now. And the language has to stay plain, so the right people see themselves in it.
The place to start is the roles that get hired most often. Pulling the hiring manager and the people who do the work into the draft, then refining from there, is what gets it right. Once the words match the work, the gains build on themselves.
For the writing itself, Ongig helps teams draft job descriptions, strip out biased wording, and manage them at scale.
