A live, cross-context intervention delivered to apprentices across the UK, US, and Canada

Qualified. Capable. But Not Being Heard.

A systems-level analysis of the structural communication gap in UK and international apprenticeship development. Grounded in pre-session survey data from a cohort of professionals at Amazon, KPMG, Google, NatWest, BBC News, UBS, SAP, Unilever, Accenture, and thirty further organisations, this report examines the failure modes of the earn-and-learn model, quantifies the gap between technical qualification and professional communication capability, and presents the Apprentis LEO Framework as a principled, evidence-grounded architecture for closing it.

Abstract

This report presents the findings of a structured pre-session survey administered to forty-one professionals prior to the Asking Better Questions LinkedIn Live session hosted by Apprentis and workforce strategist Kara Gooch. The registrant population includes individuals currently engaged in degree apprenticeships and professional roles at organisations including Amazon, KPMG, Google, NatWest Group, BBC News, UBS, SAP, Unilever, Accenture, Capgemini, Arup, and Turner & Townsend, organisations whose apprenticeship programmes represent, by any reasonable measure, the upper quartile of provision quality and employer investment in the United Kingdom.

Fifteen respondents completed the full pre-session survey instrument across fifteen distinct professional sectors. Their responses reveal a structural pattern of consistent and measurable divergence between measured professional communication skill and the confidence required to deploy that skill in the real-time conditions of workplace practice. The average confidence score of 67 percent trails the average questioning and listening skills score of 75 percent by eight percentage points. Twelve of the fifteen respondents scored their confidence in generating questions that produce actionable outcomes between four and six on a ten-point scale. Not one reached the upper band. Ninety-three percent have no consistent structured framework for professional questioning (47% Rarely, 47% Sometimes, only 7% Often - zero Always).

The argument of this report is that these findings are not incidental to the earn-and-learn model of apprenticeship development. They are produced by it. The earn-and-learn model is a system optimised to deliver two of the three dimensions that constitute a complete professional: learning and earning. It does not design for the third dimension, what the Apprentis LEO Framework denominates as Own, and consequently does not produce it. This report provides the theoretical architecture, the empirical grounding, the systems analysis, and the design specification required to correct that omission.

I. The Demand Signal: Who Registered and Why It Matters

41 Professionals. 36 Organisations. One Common Gap.

The demand signal before the data

+ 20 more organisations across media, technology professional services, and financial institutions

The forty-one professionals who registered for the Asking Better Questions session did not constitute a random sample of the apprenticeship population. They were a self-selected cohort, drawn toward a session specifically designed to address a professional capability gap that they had already identified as present in their own practice. That self-selection is not a methodological limitation. It is a finding, and arguably the most commercially significant finding in this entire report.

The organisations represented in the registration list are not organisations whose apprenticeship programmes are struggling for investment, expertise, or institutional commitment. Amazon's UK apprenticeship programme is one of the largest in the country. KPMG and Accenture have invested in degree apprenticeship infrastructure over more than a decade. UBS, Unilever, and Capgemini operate dedicated early careers functions with professional L&D capacity. Arup and Turner & Townsend are sector leaders in structured professional development for technical apprentices.

When forty-one people from organisations of this calibre register for a session on professional communication, the appropriate analytical response is not to attribute it to individual curiosity. It is to recognise it as a demand signal: evidence that the gap exists inside even the most well-resourced programmes in the system, and that the people operating within those programmes are sufficiently aware of it to seek external provision.

The registration data establishes three things simultaneously. First, the communication gap is not a problem concentrated in under-resourced or poorly designed programmes. Second, the existing provision infrastructure, even at elite employers, is not closing it. Third, there is latent market demand for a structured, evidence-based intervention and that demand is concentrated precisely in the high-budget, development-oriented employer segment that is most capable of acting on it and most commercially valuable to Apprentis.

"This is a proof of market. 41 professionals at organisations with the resources to solve this problem internally chose instead to attend a session designed to give them what their programmes had not. That is your total addressable market signaling its own demand."
Iyioluwa Adesan, Founder, Apprentis

II. Cross-Context Architecture: Why Three Countries Matter

Thirty-six of the forty-one registrants are based in the United Kingdom. Four are based in the United States, representing roles at Accenture and Accredible. One registrant is based in Canada, a Senior Business and Integration Analyst at Accenture. This cross-context distribution is not incidental to the Apprentis pilot design; it is central to the claim that the observed communication gap is structural rather than geographically specific.

The Apprentis LEO Framework argues that the Own dimension deficit is not a policy failure specific to the UK apprenticeship system, nor a cultural quirk of British professional formation, nor an artefact of the regulatory architecture of the Apprenticeship Levy. It is a structural feature of the earn-and-learn model wherever that model operates: the predictable output of a system that has defined its success criteria around technical qualification and workplace participation, and has not defined its success criteria around the professional identity, agency, and communication capability that determine whether qualification translates into sustained professional impact.

The UK apprenticeship system, the US registered apprenticeship system, and the Canadian employer-sponsored professional development model are structurally distinct. They differ in funding mechanism, regulatory oversight, assessment architecture, and cultural positioning. What they share is the earn-and-learn proposition at their core, and what this cohort's cross-context registration data suggests is that they also share the consequence of that proposition's incompleteness. The gap does not care which country you are in. It cares which model you are operating in.

Methodological note: The cross-context dimension of this pilot is directional rather than statistically controlled at this stage. The US and Canadian registrants were drawn from a single professional network and cannot be treated as representative of the broader populations of their respective apprenticeship systems. The claim being made is that the signal is present across contexts, not that its magnitude has been precisely quantified across them.

"The earn-and-learn model is global. So is the gap it produces. A framework that closes it does not require redesign for each national system. It requires one architectural correction, applied consistently: the Own dimension."
Iyioluwa Adesan, Founder, Apprentis

III. The Numbers: What the Pre-Session Data Shows

The pre-session survey instrument collected responses across four scored dimensions: questioning and listening skills, professional confidence, engagement and leadership intentions, and demographic and contextual variables. The instrument also collected open-ended qualitative responses on participants' experienced challenges and development goals. The following sections analyse each dimension in turn before integrating the findings into a systems-level account of what the data collectively demonstrates.

IV. The Systems Failure: Why the Earn-and-Learn Model Produces This Gap

Before examining the quantitative findings in detail, it is necessary to establish the theoretical framework within which they are interpreted. Data does not speak for itself. Numbers require a model to give them meaning, and the model that gives the numbers in this report their significance is the Apprentis LEO Framework, grounded in a systems-level analysis of how the earn-and-learn model was designed, what it was designed to produce, and what it structurally cannot produce without modification.

The earn-and-learn model is, in systems design terms, a two-input, two-output architecture. The inputs are occupational knowledge, delivered by training providers and universities, and workplace experience, delivered by employers. The outputs are technical qualification, evidenced through end-point assessment, and employment, evidenced through wage payment and job role performance. The system is coherent, relatively well-specified, and measured against its own defined outputs moderately effective. The problem is that its defined outputs are not the complete set of outputs required for a genuinely transformative professional education.

The third output, what the LEO Framework denominates as Own, encompasses the professional identity, communication agency, and self-directed career capacity that determine whether the first two outputs translate into sustained professional impact. This dimension is not an optional enrichment to the earn-and-learn system. It is the feedback mechanism that allows the system's other outputs to compound over time. A professionally formed individual with strong communication capability extracts more value from every subsequent learning and earning experience. A technically qualified employee without that formation extracts progressively less. The Own dimension is, in information-theoretic terms, the signal amplifier. Without it, the system's outputs attenuate rather than compound.

The absence of the Own dimension from the earn-and-learn system is not an oversight. It is a design consequence of how apprenticeship success was historically defined. The medieval apprenticeship model, from which today's earn-and-learn system descends, operated in a stable and hierarchical occupational world. Professional identity was often conferred externally through guild membership, social position, and community structures rather than intentionally developed by the education system itself. Those structures no longer exist. Today's professional environment is dynamic, less hierarchical, and continuously changing. Yet many apprenticeship systems still assume identity, confidence, and career agency will emerge naturally.

As a result, the system continues to produce technically qualified individuals who may lack the professional self-authorship needed to navigate modern careers.

"They know more than they feel able to use. The earn-and-learn model is not broken. It is incomplete. It was designed for a world where professional identity was conferred externally, and communication capability was developed informally. That world no longer exists. The LEO Framework is the system update."
Iyioluwa Adesan, Founder, Apprentis

V. Signal Extraction: What Fifteen Responses Tell Us

A responsible engagement with the limitations of this dataset is a prerequisite for making credible claims from it. Fifteen survey responses do not constitute a statistically representative sample of the UK degree apprenticeship population, which numbers in the hundreds of thousands. Confidence intervals at this sample size are wide, and any finding that depends on precise point estimates would be unreliable. This report makes no such claims.

What fifteen responses can provide, however, is something analytically distinct from statistical representativeness: signal consistency. A recurring directional pattern observed across multiple cases, particularly across different sectors, organisations, and contexts can provide meaningful evidence of an underlying structural phenomenon even when the sample size is small.

The logic is cumulative rather than absolute: each additional observation that reproduces the same pattern increases confidence that the signal is real rather than incidental. Across fifteen respondents spanning fifteen employers, fifteen sectors, multiple programme types, and three national contexts, the repeated direction of the findings merits serious attention.

On sample provenance: All fifteen survey respondents were drawn from the pool of forty-one registrants. Registration was voluntary and driven by personal identification with the session's stated focus on professional communication development. This selection condition means the cohort is biased toward individuals who are already aware of a development need, which may mean the confidence gap in the broader apprenticeship population is larger, not smaller, than what this cohort shows. Self-selecting toward awareness of a gap does not inflate the gap; it may, if anything, suppress it.

"Fifteen responses. Fifteen different employers. Fifteen different sectors. Three countries. The same direction of gap in every case. At some point, the appropriate statistical response to a consistent signal is not to ask for a larger sample. It is to ask why the system keeps producing it. If the gap exists inside Amazon, KPMG, and Google's apprenticeship programmes, it exists everywhere the earn-and-learn model operates without the Own dimension."
Iyioluwa Adesan, Founder, Apprentis

VI. The Confidence Deficit: Anatomy of an Eight-Point Gap

6.1 The Primary Finding

The average questioning and listening skills score across all fifteen participants is 75 percent. The average confidence score is 67 percent. The eight-point mean divergence between these two dimensions is the primary quantitative finding of this report, and it is significant not merely because of its magnitude but because of its direction and its consistency.

The direction is invariant: in ten of the fifteen cases, confidence trails skills. This is not a random distribution of divergences in both directions. It is a systematic pattern of capability outpacing confidence of apprentices who can construct effective questions but who lack the professional assurance to deploy them in the conditions where it matters. This directional consistency is itself analytically important. It rules out the hypothesis that the gap is driven by measurement error in the skills dimension and confirms that it reflects a genuine structural divergence between the latent capability and the enacted performance of these professionals.

6.2 The Technology Sector Pattern

The three technology-adjacent participants in this cohort, working in financial technology, computer software, and IT infrastructure, show a collective average confidence score of 47 percent, compared with 72 percent for the remaining twelve participants. The 25-point differential is the sharpest sector-based observation in the dataset. It is drawn from three individuals, not a sector-representative sample, and should be interpreted accordingly: as a directional signal warranting further investigation rather than a statistically precise sector estimate.

The signal is nevertheless consistent with what first principles might suggest. Technology-adjacent professional environments are often characterised by high cognitive load in technical domains, asynchronous communication norms, text-mediated collaboration, and interaction patterns that reward precision over frequency. In practice, these can provide fewer natural opportunities to build confidence in real-time spoken professional communication.

The technology confidence gap is 25 points. In the industries building the next decade of economic value, the apprentices entering them cannot deploy the skills they demonstrably have. That is not a talent scarcity problem. It is a programme design specification written incorrectly.
Iyioluwa Adesan, Founder, Apprentis

VII. Self-Report and Measured Scores: The Phenomenology of the Gap

The convergence is strong. Forty-seven percent of participants, the largest single group, describe themselves as 'somewhat confident.' This formulation deserves careful analytical attention. Somewhat confident does not denote a lack of awareness or motivation. It denotes a conditional, context-sensitive, qualified self-assessment: the sense that one's communication capability is adequate in some conditions but not reliably so across the range of conditions that professional practice generates. This is the phenomenology of the confidence gap: not an absence of capability but an absence of the assurance that capability will be accessible when it is most needed.

The zero finding, not one of fifteen degree-level apprentices at elite employers scoring their actionable confidence above eight out of ten is the single most consequential number in this dataset. It is not a finding about the lower tail of the distribution. It is a finding about the entire distribution. The upper band of actionable communication confidence is empty. In a cohort drawn from Amazon, KPMG, Unilever, MBDA, Savills, and the Royal Devon NHS, operating in programmes with significant employer investment and professional development infrastructure, the upper band is empty. The interpretation is not that these individuals lack capability. The interpretation is that the system that developed them did not develop this.

Zero in the upper band. Not a tail finding. A distributional finding. The entire cohort, at elite employers with full programme investment, is concentrated in the moderate range. The question is not whether the system is failing this dimension. The question is whether anyone in the system is measuring it. Now ask yourself what your performance review process is measuring.
Iyioluwa Adesan, Founder, Apprentis

VIII. The Framework Deficit: A Provision Failure With a Known Remedy

The finding that 87 percent of participants use no consistent framework for professional questioning is one of the most actionable insights in this report. It points to a specific and remediable gap rather than a vague confidence issue.

These participants are not lacking motivation or potential. They are often entering important workplace conversations without a repeatable method for asking strong questions, following up effectively, or extracting useful insight.

A questioning framework functions as a professional scaffold. It helps individuals identify what information matters, how to ask for it clearly, and how to build on responses in real time.

Without that scaffold, apprentices may still gain value from mentoring sessions, feedback conversations, stakeholder meetings, and manager check-ins but often less value than those interactions could provide.

87% have no consistent framework. They are not absent through indifference. They are absent because the system that designed these programmes did not identify their provision as a requirement. What is not specified is not built. What is not built is not measured. What is not measured does not create pressure for change. This is the self-reinforcing feedback loop that keeps the gap open.
Iyioluwa Adesan, Founder, Apprentis

IX. Qualitative Phenomenology: Seven Failure Modes of Professional Communication

The most analytically significant theme, appearing in three of the fifteen responses in various formulations, is what might be called pressure-induced communication collapse: the experience of communication capability that is present in reflective or low-stakes conditions but that degrades under the real-time pressure of professional observation. This is a specific and diagnostically important failure mode because it is precisely the mode that the earn-and-learn model's assessment architecture is least equipped to detect.

Mo Maygi, a Data Science Degree Apprentice at Unilever, named it simply: 'Not getting stuck over my words.'

That phrase surfaces the same experience that three participants described in different formulations across the dataset: communication capability that is present in reflective conditions but that degrades at precisely the moment it matters most, when being observed, assessed, or required to respond in real time.

Capability that cannot be expressed is invisible talent. Invisible talent produces misjudged performance reviews, missed promotions, and professionals who leave because they were never fully seen. This is not a soft problem. It is a measurable, consequential, and remediable structural failure.
Iyioluwa Adesan, Founder, Apprentis

XII. The Equity Dimension: Who Bears the Cost of the Gap

The demographic composition of this cohort, 60 percent identifying as Black, African, Caribbean, or Black British; 20 percent as Asian or Asian British; 20 percent as White British, Irish, or Other is not representative of the UK degree apprenticeship population nationally, in which white apprentices continue to form the majority of starts at higher levels. This over-representation of ethnic minority professionals in a cohort drawn toward a communication development session is not incidental. It reflects a pattern in the research literature on professional identity formation, workplace communication, and career mobility that is sufficiently consistent to warrant direct engagement.

For these professionals, the confidence gap that this cohort's data reveals is not merely a development gap. It is an equity gap: the additional friction between capability and deployment that arises from the need to communicate not only effectively but across a cultural translation layer that their white colleagues do not face.

The communication gap falls hardest on the professionals who are already carrying the greatest cognitive and social load. For 80 percent of this cohort, closing the gap is not merely a professional development objective. It is a structural correction to an uneven playing field that their programmes did not acknowledge and did not address.
Iyioluwa Adesan, Founder, Apprentis

XIII. The Leadership Signal: What 73 Percent Tells Us About Latent Potential

When asked whether they would be interested in joining a regional apprentice community, not one of the fifteen participants said no. Sixty-seven percent said yes outright, and thirty-three percent indicated openness. When asked whether they would like to explore a leadership role in their region, seventy-three percent responded affirmatively or with qualified openness.

These are the same fifteen individuals who averaged 67 percent on professional confidence. The aspiration to lead, to contribute, and to build community is not dampened by the confidence gap. It exists independently and at a higher magnitude than the confidence dimension would, in isolation, predict.

73% want to lead. The pipeline is not empty. It is full of capability that the system has not yet equipped to be visible. The organisation that closes that gap does not merely develop better apprentices. It earns the compounding return of professionals who contribute back to the ecosystem that shaped them.
Iyioluwa Adesan, Founder, Apprentis

Published: April 2026 | Registrants: 41 professionals from 36 organisations | Survey: 15 apprentices across 15 sectors | Geography: UK (primary), US, Canada | Framework: LEO (Learn, Earn, Own) | Apprentis | apprentisapp.com