The Experience Paradox: Why It's Getting Worse

The requirement for entry-level jobs has gotten worse. Not marginally. Measurably.

A 2025 Harvard Business School analysis of 50 million job postings found that the share of entry-level positions explicitly requiring three or more years of experience rose by 18 percentage points over the preceding five years. That is not a trend buried in a footnote. It is a structural shift happening in plain sight.

We first wrote about the experience paradox several years ago — the condition where roles labelled “entry-level” demand professional experience. Since then, the problem has intensified rather than softened.

This post is a survey of what has been tried, why those solutions fell short structurally, and what a durable fix actually requires.

What Was Tried: Three Main Approaches

Internship Programs

The most established response to the experience gap has been the internship. Companies offer short-term placements, candidates gain exposure, and ideally the relationship converts into a full-time offer.

The problem with this approach is not the concept. It is the scale and the selection bias. Competitive internship programmes now receive thousands of applications for hundreds of spots. The candidates who land these placements tend to be students at well-connected institutions. The internship market has become a secondary filter, not a correction mechanism.

Beyond selection bias, the internship-to-full-time conversion rate is lower than many assume. A 2024 NBER working paper tracked conversion rates across major US employers and found fewer than 40% of interns at mid-to-large tech firms received a full-time offer within 18 months of completion. The rest returned to the same application pool, now with experience the market did not value as a signal.

The structural issue is that internship pipelines select for candidates who would likely have navigated conventional hiring successfully anyway. The candidates excluded from competitive internship markets are precisely the ones the experience paradox hits hardest.

Skills-Based Hiring Initiatives

In 2022, Google, Apple, IBM, and a cohort of large employers announced revisions to hiring practices that would deprioritise degree requirements in favour of verified skills. The announcement was significant. The execution was narrower than the coverage suggested.

Google’s Career Certificates programme targets mid-career pivoters and entry-level adjacent candidates with a credential pathway into associate-level roles. The programme has produced outcomes. But when researchers at Harvard’s Labor Market and Skill Formation project tracked placement rates for certificate holders into genuinely competitive roles, they found that the conversion to roles with advancement potential was heavily concentrated in a narrow band of the market — individuals who already had some prior professional background, even if non-degree-holding.

Apple’s Skills Training initiative has operated on a similar pattern. The hiring exceptions — roles where degree and credential barriers genuinely came down — exist primarily in enterprise customer support, retail operations, and supply chain associate roles. In software engineering, product management, and data science roles at the same company, the median qualified hire still comes from a conventional degree or prior large-company background.

IBM’s SkillsBuild initiative has perhaps the most documented internal implementation. Their internal data shows a measurable increase in hires from non-traditional backgrounds for certain role categories between 2022 and 2024. But the same internal data shows that the effect is concentrated in early-career customer success, technical support, and associate analyst roles. In mid-seniority product and engineering roles, the skills-based pathway accounts for a low single-digit percentage of hires.

The gap between the announcement and the outcome is not dishonesty. It is a structural observation: large established companies redesign hiring at the margins because the margins are where they can absorb risk. Core roles remain protected. The signal sent by a high-profile skills-based hiring announcement is real in terms of intent; it is limited in terms of market-wide effect because the roles where credential barriers most impede entry are not the roles being restructured.

Coding Bootcamps

The bootcamp model arrived with genuine promise. A condensed, intensive credential that demonstrated applied skill, with hiring outcomes measured and published. For a period it worked.

What happened next is instructive. As bootcamp graduation rates climbed and hiring volume from bootcamp pipelines increased, companies that had been hiring bootcamp graduates began receiving a high volume of applicants with the same credential and roughly similar project portfolios. The filter shifted from “no degree” to “which bootcamp, which projects, which outcomes.”

A credential that becomes a filter behaves identically to a degree requirement. It just has a different label. The bootcamp’s original value proposition was that it was a pathway that bypassed credential gatekeeping. As it scaled, it became another credential gate — with its own internal hierarchy and proxy signals that employers learned to read.

Why These Solutions Fell Short Structurally

The common thread in all three approaches is that they address the symptom rather than the mechanism.

Internships do not solve the experience paradox. They create a smaller, more competitive market for a subset of experience signals. Skills-based hiring at large companies does not solve the paradox. It creates a carve-out for a limited number of roles while leaving the core hiring infrastructure unchanged. Bootcamps do not solve the paradox. They create a new credential class that stratifies rather than replaces the existing filter.

In each case, the employer retains the ability to add layers of evaluation beyond the stated requirement. An internship requirement becomes a baseline. A skills-based credential becomes a minimum. The target moves because the underlying incentive has not changed: employers face lower cost from a rejected qualified candidate than from a hired unqualified one. In the absence of structural changes to that cost equation, filters persist and evolve.

The experience paradox is not a communication problem. It is an information asymmetry problem. Employers cannot cheaply distinguish between candidates who will perform well and candidates who will not, so they use proxies — credentials, prior roles, specific tool familiarity. Those proxies are imperfect but low-cost. The solutions that have been tried mostly try to offer better proxies. None have tried to change the underlying information problem.

This matters because every solution that works within the existing evaluation structure is temporarily effective at best. As soon as a solution gains adoption, it becomes part of the filter set. The experience paradox does not disappear — it relocates.

Delivery Records as the Structural Fix

This is where delivery records differ from every approach above.

A delivery record is a documented, verifiable account of work completed. Not an internship experience. Not a course completion. Not a portfolio project with curated outcomes. A real piece of work, produced under real conditions, with a real output that can be checked.

The structural advantage is this: it shifts the information asymmetry. When an employer can see a candidate’s actual output — a report written, a dataset analysed, a system built, a campaign delivered — the proxy evaluation collapses. The employer does not need to infer potential from credential or background. They can read the work.

This is not a theoretical framing. FursaFlow operates on this principle. Every learner builds a delivery record across real projects — work you can show to any employer, not just list on a CV. The work exists independently of any label attached to it. An employer reviewing a delivery record is not evaluating a bootcamp credential or a degree status. They are reading evidence.

For more on how delivery records work in practice, see our guide: How to Build a Portfolio with No Experience.

The structural difference from all prior approaches is in the verification mechanism. Credentials and portfolios can be inflated. Delivery records from a system with auditable output are harder to misrepresent because the work itself is the signal. The candidate cannot claim competence in an area they have not delivered in. The work is the claim.

There is a second structural difference worth noting. Delivery records are not portable in the same way a credential is. You cannot simply add them to a resume in a formatted section and move on. They require the employer to engage with the content, not just the label. That friction is the feature. It is what changes the hiring interaction from a credential evaluation to a work evaluation.

The experience paradox persists because the market has not had a reliable mechanism for evaluating candidates without defaulting to credential proxies. Delivery records are not a better proxy. They are a different information structure — one where the output is the evaluation, not a stand-in for it.

If you are early in your career and tired of being filtered out of roles you could do, this is the infrastructure designed around your situation. Not another credential. Not another internship lottery. Work that speaks for itself.

Be Part of the Solution