Key findings
- England’s new apprenticeship assessment arrangements are a redistribution of responsibility, not a simple removal of quality controls: assessment organisations and regulators retain formal responsibilities for validity, consistency and quality assurance.
- Employers must now verify occupational behaviours for reformed standards at gateway to completion, making line-management practice an operational part of the apprenticeship system.
- The immediate risk is uneven implementation. Large employers may be able to formalise supervision and evidence; smaller firms may need proportionate support rather than extra paperwork.
- Starts are rising, but the age profile matters: in August 2025 to April 2026, people aged under 19 represented 20.6% of apprenticeship starts in England. Reform should therefore be judged by access and progression as well as volume.
- Sanctuary’s recommendation is a short employer capability compact: define observable behaviours, assign a trained workplace lead, establish monthly developmental reviews, maintain a light evidence trail, and use provider challenge before gateway.
The policy change is live; the operational question is only beginning
England’s 2026–27 apprenticeship funding rules apply to apprenticeships starting on or after **1 August 2026**. They use the term “growth and skills levy” for the apprenticeship levy being transformed by government, while revised assessment plans are being introduced in stages rather than through a single national switch-over (Department for Work and Pensions, 2026) ↗.
The assessment reforms aim to reduce duplication, permit more flexible timing and methods where appropriate, and allow providers to deliver and mark some assessment elements. That does **not** mean external assurance has disappeared: assessment organisations remain responsible for developing assessment approaches and, where providers assess, for quality assurance; Ofqual and the Office for Students retain regulatory roles in their respective parts of the system (Department for Work and Pensions, 2026) ↗.
The practical shift is sharper for employers. For revised plans, employers verify whether the apprentice has demonstrated the behaviours in the occupational standard before gateway to completion. Skills England’s July guidance says this should be based on naturally occurring workplace evidence — supervision, feedback, induction conversations and regular check-ins — rather than a new dossier of paperwork (Skills England, 2026) ↗.
**Interpretation:** this is best understood as a transfer of a bounded but consequential judgement into routine management. An employer that cannot give specific feedback, distinguish observed conduct from personal affinity, or create opportunities for the apprentice to demonstrate the standard will struggle to make that judgement consistently. The issue is not whether a manager can complete a checklist; it is whether the workplace actually functions as a reliable learning environment.
Why this matters now: volume is growing, but access and quality cannot be assumed
Provisional official statistics record **308,770** apprenticeship starts in England from August 2025 to April 2026, an 8.7% increase on the equivalent period a year earlier. Participation rose to 735,500 and achievements to 146,210. Yet under-19s accounted for 63,530 starts, or **20.6%** of the total, while people aged 25 and over accounted for 52.4% (Department for Education and Department for Work and Pensions, 2026) ↗.
These figures do not establish that assessment reform caused any change in participation; the reforms are still being phased in and the statistics are provisional. They do, however, make clear why implementation quality matters. Apprenticeships are being asked to serve workforce development, occupational progression and entry routes for younger people at the same time. A system that becomes easier for established employees in well-resourced organisations but harder for smaller employers or new entrants to navigate would meet only part of that mission.
Small and medium-sized enterprises are especially material. In the latest employer-size release, **37%** of apprenticeship starts in 2023/24 were with SMEs, while most employers with at least one apprenticeship start were small. That combination suggests that many participating employers are likely to have limited specialist HR, learning-and-development or compliance capacity (Department for Education and Department for Work and Pensions, 2026) ↗.
**Interpretation:** the reform’s promise of proportionality is plausible only if the operating model is proportionate too. Replacing duplicate end-point processes with opaque, inconsistent manager judgement would not reduce burden; it would relocate risk onto apprentices, line managers and small firms.
What the evidence says about workplace capability
Research on degree apprenticeships cannot be mechanically generalised to every occupational standard, but it offers a useful warning. A recent UK study of computing degree-apprenticeship mentors found that workplace mentors play a central role in guiding apprentices, facilitating suitable work opportunities and supporting tripartite relationships. In its mentor survey, most respondents had not received training from the university; the authors identify a need for more purposeful support, including around inclusion (Taylor-Smith et al., 2026) ↗.
Earlier qualitative research likewise found that workplace mentoring can benefit apprentices, mentors and organisations, while stressing that effective mentoring requires intentional practice rather than merely allocating a colleague to the role (Mulkeen et al., 2019) ↗. These are small-scale studies, not national causal evaluations. Their value here is diagnostic: workplace learning depends on structured opportunities, informed support and relationships that make development visible.
This aligns with the design of the English reforms. Skills England explicitly states that behaviour verification should already be integral to performance management and staff development. But “should already be” is an implementation assumption, not evidence that every employer has consistent review routines, calibrated managers or psychologically safe feedback channels (Skills England, 2026) ↗.
A credible employer judgement should therefore meet four tests:
1. **Observable:** the behaviour is translated into work situations a manager can genuinely see.
2. **Developmental:** feedback occurs early enough for an apprentice to improve, not only at gateway.
3. **Calibrated:** managers know what the standard means and when to seek provider or assessment-organisation advice.
4. **Fair:** evidence is drawn from repeated work activity, with reasonable adjustments and inclusion considered rather than assumed.
These are Sanctuary’s analytical criteria, not additional regulatory requirements.
Sanctuary recommendation: build a five-part employer capability compact
For employers, providers and local partnership organisations, the priority is to make the new responsibility workable without recreating unnecessary bureaucracy.
**1. Start with a behaviour-to-work map.** Before the apprentice begins, convert each required behaviour into two or three normal workplace contexts in which it can be observed. “Professionalism”, for example, is too abstract until connected to client handovers, safety briefings, team problem-solving or response to feedback.
**2. Name one accountable workplace lead.** The line manager may delegate coaching, but responsibility for the final employer confirmation should not be dispersed across informal supervisors. The named lead needs time, a clear escalation route and a basic understanding of the relevant standard.
**3. Hold short monthly development reviews.** A 20-minute structured conversation can record examples, identify missing opportunities and agree the next work task. This is more useful than reconstructing evidence shortly before gateway.
**4. Create a light, auditable trail.** Retain concise records of observed examples, feedback, agreed actions and reasonable adjustments. The purpose is developmental continuity and defensibility — not an evidence archive for its own sake.
**5. Use tripartite challenge before gateway.** Provider, apprentice and employer should test whether the workplace evidence is sufficient and whether there are gaps in exposure. Assessment organisations remain responsible for assessment design and quality assurance; employers should seek clarification early where a judgement is uncertain (Department for Work and Pensions, 2026) ↗.
For SMEs, this compact should be offered as a shared local service: a template, a short manager briefing, peer support and access to an adviser. The most effective intervention may not be another national campaign to promote apprenticeships. It may be practical assistance that helps a busy supervisor turn everyday work into credible learning and fair evidence.
This complements Sanctuary’s earlier argument that employer capability is now the quality question in apprenticeship reform: assessment reform has moved more of the quality challenge into the workplace. It also creates a practical role for Educational Support Services and Human Resources Support where organisations need to align learning, supervision and people-management practice.
The test for the next year
The reforms should not be judged solely by whether assessment becomes faster or less burdensome. Those are legitimate aims, but an apprenticeship is a route into occupational competence and progression. The stronger test is whether apprentices receive more timely feedback, employers can make sound judgements with confidence, and smaller organisations can participate without being overwhelmed.
There is a credible upside: assessment that is closer to real work can reduce artificial duplication and make the apprenticeship feel less like a programme bolted onto a job. There is also a credible risk: without manager capability, inconsistency may be hidden behind the language of flexibility.
**Sanctuary perspective:** England should treat line-manager capability as enabling infrastructure for apprenticeship quality. The policy has made this infrastructure more visible. Local skills systems, providers and employers now need to build it deliberately.
Research foundation
References
- Department for Work and Pensions (2026). Changes to apprenticeship assessment: 2025 to 2026. GOV.UK.Source ↗
- Department for Work and Pensions (2026). Changes to apprenticeship assessment: 2025 to 2026. GOV.UK.Source ↗
- Skills England (2026). Apprenticeship behaviour verification guidance for employers. GOV.UK.Source ↗
- Department for Education and Department for Work and Pensions (2026). Apprenticeships: academic year 2025/26. Explore education statistics.Source ↗
- Department for Education and Department for Work and Pensions (2026). Apprenticeships in England by industry characteristics: academic year 2023/24. Explore education statistics.Source ↗
- Ella Taylor-Smith, Sally Smith, Matthew Barr, Maria Cecil and Andrea Scott (2026). What training do workplace mentors receive to support degree apprentices? An activity theory analysis, focusing on inclusion. Journal of Education and Work, 38(5–8), 165–187.Source ↗DOI: 10.1080/13639080.2026.2619947
- John Mulkeen, David Abdou and David Leigh (2019). Workplace mentoring of degree apprentices: developing principles for practice. Higher Education, Skills and Work-Based Learning, 9(2), 211–224.Source ↗DOI: 10.1108/HESWBL-10-2018-0108
- Sanctuary Consulting & Development Group. Hero image: administrator-supplied photograph. Owner supplied / permission confirmed.Image source ↗
Related Sanctuary capabilities
From analysis to implementation.
Discussion

No approved comments yet.