How do you get your first chance—or your next one—when hiring systems only recognize what you’ve already done?
We keep telling people to prepare for a changing world of work. Learn new skills. Go back to school. Find a career that has a future.
But what happens if they do all of that, and the systems connecting them to their next opportunity still understand them primarily through the jobs they used to have?
That question has stayed with me since my conversation with Matt Plummer on The Signal Lab.
Matt is Chief Product Officer at Appcast and spent twelve years at ZipRecruiter. We’ve worked together, challenged each other, and thought about marketplaces for nearly twenty years. In this interview, I asked him what a marketplace should learn about someone to recommend the right jobs. Suppose a job seeker gives you fifteen minutes. What do you ask?
Matt’s starting point was practical: “What jobs have you had in the past?”
Understand the work someone has done, infer the skills they probably needed to do it, and use that information to identify other jobs they could do. He also named licenses and certifications. But as I kept pressing on how these systems understand capability, the conversation kept returning to work history.
There is a reason for that. A job title and a few years of experience give a marketplace something familiar to work with. A person’s entire educational experience is much harder to interpret.
Still, even Matt flagged an assumption inside that approach: “We’re also assuming people were successful in those roles.”
Having held a job is evidence. What it proves, and how far that evidence travels, are separate questions.
So I asked about the people this approach struggles to see. What about someone who doesn’t have relevant work experience yet? What about someone who has worked hard to become capable of something their employment history doesn’t reflect?
Matt recognized the first problem immediately. His niece was in a master’s program, looking for her first job in her chosen field. She had taken classes, completed projects, and demonstrated proficiency. What she didn’t have was the work history that would make those capabilities easy for a marketplace to recognize.
He described the gap clearly. A job board cannot understand every class at every university and what it took to pass it. Something has to translate that learning into evidence the marketplace can use.
The second person—the adult learning something new to change careers—didn’t get the same answer. Matt focused on the graduate. The career changer’s problem remained unresolved.
Imagine someone working retail during the day and studying cybersecurity at night. Their work history tells a clear story about retail. Their recent learning tells a different story about what they might do next. If the matching system gives most of its attention to the first story, how does the second one ever become visible?
We are asking people to change faster than our systems can recognize that they have changed.
I’ve been thinking about this at the very beginning of working life, too. For generations, places like McDonald’s have helped people get started. The company says one in eight Americans has worked at one of its restaurants. Those jobs provide a paycheck, but they also give a young person something consequential: a first chance to demonstrate reliability, learn under pressure, and build a record someone else will recognize.
When I walk into fast food restaurants today, I often notice adults doing work I associate with those first teenage jobs. My observation isn’t a national trend line. But Bureau of Labor Statistics data for 2025 show that roughly a third of fast food and counter workers were 25 or older. These jobs serve people at very different stages of life.
That doesn’t establish that older workers are displacing teenagers. It does challenge the idea that there is a separate, dependable pool of starter jobs waiting for each new generation. A teenager seeking a first opportunity and an adult supporting a household may be looking at the same opening.
Now extend the question to the first professional job. What happens if employers need fewer beginners because AI can perform some of the tasks through which beginners used to learn?
There are already troubling signals. In its August 2026 update to Canaries in the Coal Mine, Stanford’s Digital Economy Lab reported that employment among workers ages 22–25 in AI-exposed occupations was 19% below where it would have been had it kept pace with less-exposed peers. The gap operated primarily through reduced hiring. The researchers describe these as early, descriptive indicators, not proof of causation, and find no widespread, economy-wide displacement.
Even with those qualifications, the question is urgent. If there are fewer opportunities to acquire experience, relying on experience to allocate the remaining opportunities becomes a serious problem.
And what happens further along in a career, when AI changes the economic value of work someone has spent years doing?
Their capabilities don’t simply disappear. A customer service professional may have developed judgment, conflict resolution, and the ability to navigate complicated systems. An administrative professional may know how to coordinate people, manage exceptions, and keep an operation moving. Those capabilities may matter elsewhere. But a system that mostly recognizes their former job title may struggle to make the connection.
The risk is that people lose the work their experience points toward while remaining invisible to other work they could learn to do—or already have the capability to do.
That’s where I see a crisis taking shape. We could invest heavily in helping people adapt and still leave them trapped behind the same old description of who they are.
This is why the conversation with Matt eventually turned to the Comprehensive Learner Record, or CLR. 1EdTech’s standard supports portable, verifiable records of learning and achievement, including courses, competencies, skills, and workplace accomplishments.
For a student, that should mean leaving school with more than a degree title and a list of courses. A useful CLR should connect credentials to the capabilities demonstrated along the way: the research completed, the project delivered, the technical problem solved, the assessment passed. It should help someone understand the substance of an education, including documented learning beyond the classroom.
For the adult going to night school, verifiable credentials should provide evidence of newly developed capabilities while their employment history still reflects their previous career. A credential should make clear who assessed the learning, what was demonstrated, and what evidence supports the claim. The person shouldn’t have to wait for a new employer to take a chance before their new capabilities become legible.
For someone whose work is being automated, we need ways to document the capabilities that can transfer. Employer attestations, assessed work samples, professional credentials, and targeted assessments could help establish what a person can bring to another role. Some capabilities will transfer directly. Others will require additional learning. Better evidence should help distinguish between them.
None of these records captures a person’s full potential. But they can give hiring systems a better basis for recognizing capability than a job title alone.
Matt saw the promise. He also named the conditions: the information has to be available at meaningful scale, and marketplaces have to connect it to what employers actually need. Issuing a credential is only part of the work. It has to reach the systems making recommendations and change which opportunities someone gets to see.
That’s the test I want us to focus on. Can a student become visible for work they’ve learned to do? Can the night-school learner be recognized for a capability their current employer has never used? Can someone leaving an automated occupation demonstrate value beyond the title they are leaving behind?
Work experience will continue to matter. But the future of work demands that people be able to become something their work history doesn’t yet describe. Our hiring systems need to recognize that possibility—and give it a fair chance.
Listen to my conversation with Matt Plummer in Episode 8 of The Signal Lab, “How Job Boards Decide.” Find The Signal Lab on Apple Podcasts or Spotify, or listen to the episode on RSS.com.
Keep listening for the signal.
Leave a Reply