Proof | Experienced Training Ltd

PROOF OF NECESSITY

Your defence to procurement: why standard training contracts structurally cannot deliver competence transfer in regulated UK sectors

01
ADOPTION

The System Works. Staff Don't Use It.

£10bn NHS waste · 70% implementation failure

Post-implementation adoption failure is the single largest category of IT project waste. The technology went live. The budget burned. Three months later, nobody's using it.

▼ Click for evidence

NHS National Programme for IT: £12.7bn spent, £2.6bn benefits delivered, £10bn net destruction. Public Accounts Committee: "one of the worst and most expensive contracting fiascos in history." Root cause: not technology failure but adoption failure—inadequate training, no user involvement.

The pattern repeats: TSB migration (£330m + £49m FCA fine, 1.9m customers locked out), Universal Credit (£34m write-offs, 38% vs 90% expected adoption), FiReControl (£469m, zero delivery).

Your risk: 70% of software implementations fail due to poor user adoption. Only 26% of employees actually use their company's ERP. Training happens week 48 of a 52-week project—then becomes the scapegoat when staff won't touch the system.

What We Deploy Instead

48-hour critical path workflow deployment. Staff processing transactions by Monday. Burn stops when adoption starts, not when training "completes."

Sources: NAO (2025), Campion-Awwad et al. (2014), Whatfix (2024), McKinsey, Gartner
⚠️
02
BRAIN

Five Consecutive Days Violates Neuroscience

66% forgotten in 24 hours · 4-chunk working memory

The 5-day classroom model violates three foundational principles of cognitive science—each established through decades of experimental research, none seriously disputed.

▼ Click for evidence

Working memory cannot absorb it: Working memory holds 4±1 chunks (Cowan, 2001), maximum 20 seconds for novel information, processes only 2-4 elements concurrently (Sweller, 1988). Enterprise software with 10-20+ interacting elements overwhelms working memory by 3-5x. Information exceeding capacity does not encode—it's lost before learning begins.

The forgetting curve: Ebbinghaus (1885), replicated by Murre & Dros (2015)—retention drops to 34% at 1 day, 21% at 31 days. By Day 5 of consecutive delivery, Days 1-3 content has undergone massive, predictable decay with no reinforcement.

Transfer failure: Only 15% of learners successfully apply training on the job (Brinkerhoff & Mooney, 2008). 65% revert to old behaviours within 30 days. Result: 80-85% scrap learning. Spaced repetition delivers 200% better retention than massed practice—but five consecutive days is massed by definition.

What We Deploy Instead

Offload memory to machine. Copilot Agents with saved settings. Checklists not courses. Spaced reinforcement at cognitive intervals, not calendar convenience.

Sources: Sweller (Cognitive Load Theory), Ebbinghaus/Murre & Dros (Forgetting Curve), Blume et al. (89 studies)
🧠
03
BOTTLENECK

Your Training is Stuck in Three People's Heads

43-716 hours development per training hour

Theory of Constraints (Goldratt, 1984): every system's throughput is limited by its tightest bottleneck. In training content development, that bottleneck is the Subject Matter Expert.

▼ Click for evidence

Development time requirements: Chapman (2010), 249 organisations, 3,947 L&D professionals: Instructor-led training requires 43 hours development per 1 finished hour (range 22-82). Interactive e-learning: 184 hours. Advanced: 716 hours. A 5-day course needs ~1,500 development hours at average ILT ratio.

The SME reality: They treat training as secondary to operational duties, spend 25-40% of time on knowledge-sharing, average 3.8-day review wait with 23-item queues. One quantified case: 152 person-days monthly delay across 40-person team = $1.46m annual productivity loss.

Goldratt's principle: "An hour lost at a bottleneck is an hour lost out of the entire system. An hour saved at a non-bottleneck is worthless." Adding more designers or LMS admins without addressing the SME constraint is local optimisation that cannot improve throughput.

What We Deploy Instead

AI agents extract SME knowledge in 60-min interviews. Capture workflows and edge cases without disrupting day. Content built in days not months. The constraint is addressed, not scaled around.

Sources: Chapman Alliance (2010), Goldratt (Theory of Constraints), systematic literature review (91 studies)
⛓️
04
INCENTIVES

Good Trainers, Wrong Contract

90% of contracts reward duration not delivery

Steven Kerr (1975): "On the Folly of Rewarding A, While Hoping for B"—organisations systematically reward behaviours they claim to discourage.

▼ Click for evidence

The folly: Rewarded (A): Days delivered, heads trained, courses completed, hours billed. Hoped for (B): Competence achieved, behaviour changed, business outcomes improved.

T&M reality: 90-95% of outsourcing operates on T&M (Genpact). Linear relationship between time spent and revenue. Zero incentive for efficiency. Reducing project duration reduces income. Providers benefit from scope creep, extended timelines, client dependency. This is rational response to incentive structure.

The Cobra Effect: British Delhi offered cobra bounties. Locals bred cobras for income. Bounty cancelled, breeders released snakes—worsening the original problem. In training: paying per course day incentivises more course days, not better courses.

What We Contract Instead

Fixed process delivery. "Staff will process X transactions/day using Y workflow by Week Z." Paid to deliver intervention, not run courses. Outcome contracted with process owners before training starts.

Sources: Kerr (Academy of Management Journal, 1975), NAO Payment by Results (2015), Genpact
🎯
05
VAGUE

"Make Them Confident" Isn't an Outcome

50% measure only satisfaction · 3% measure business results

Kirkpatrick Model (1959): training has four evaluation levels. The data on how organisations actually evaluate is damning.

▼ Click for evidence

How organisations evaluate: 50% at Level 1 only (satisfaction "smile sheets"), 34% reach Level 2 (knowledge tests), 13% reach Level 3 (behaviour change), Only 3% evaluate Level 4 (business results). Result: Only 11% of employees can fully use their training in their job.

Mager's framework (1962): Over 4 million copies sold. Effective objectives require observable performance, conditions, and criterion. Rejected vague verbs: "understand," "know," "appreciate," "be familiar with."

The contrast: "Participants will be familiar with the billing process" (unmeasurable) vs "Using the CRM system, participants will process minimum 12 customer billing adjustments per hour with <2% errors" (testable).

What We Specify Instead

Observable performance under stated conditions to stated criteria. "Process 100 invoices/day, zero errors, using system X by Week Y." Flyvbjerg's Reference Class Forecasting requires measurable outcomes—vague specs make failure invisible.

Sources: Kirkpatrick (1959), Mager (1962), 24×7 Learning (2015), NAO (2021), Flyvbjerg
📊
06
COMPLIANCE

Tick-Box Theatre vs Actual Competence

40% of FCA firms have training gaps · £81k personal fine

In UK regulated sectors, the gap between training completion and behaviour change is material regulatory risk.

▼ Click for evidence

The evidence: FCA spot checks (2023): 40% of firms had gaps in training records. Gallup: Only 10% of employees report compliance training impacted work practices. NHS: Only 8% of Trusts aligned to most recent Core Skills Training Framework.

Regulatory exposure: FCA Consumer Duty (Principle 12) requires firms demonstrate delivery of good outcomes—raising the bar beyond tick-box completion. Senior Managers & Certification Regime means personal liability.

Real consequences: TSB's former CIO personally fined £81,620 for conduct rule failures during 2018 migration disaster. Not corporate penalty—personal.

What We Evidence Instead

Process competence documentation. Staff can actually do it. Outcomes are improving. Evidence for regulators who don't accept "we delivered the training" as defence. FCA/CQC/Ofgem measure business outcomes, not course completion certificates.

Sources: FCA spot checks (2023), Gallup, HEE/Skills for Health, FCA SMCR enforcement
⚖️

Stop the Bleed

The evidence is overwhelming. The standard model doesn't work. Your procurement decision needs defending.

Surgical intervention stops the burn before it compounds.