The Context
An organisation was actively redesigning its main learning product — a large catalogue of professional development and soft skills topics — transitioning from long-form pre-recorded courses into short-form live sessions. My role was leading the instructional design process, overseeing content architecture, quality standards, and the production workflow.
The ID team sat in the middle of a five-part delivery chain, dependent on SME inputs upstream and directly responsible for the facilitator experience downstream.
Leadership
→
SME
→
ID team
→
Facilitators
→
Learners
When leadership mandated a full catalogue iteration in a tight three-month window, I used the iceberg model as a diagnostic lens — to surface patterns to anticipate and uncover what needed to change.
Guiding question
What does it take to design for skills application in real-life workplace contexts — inside a live, high-volume production system?
The Iceberg Model Diagnosis
Click each layer to expand. The iceberg reveals what was visible, what was recurring, and what needed to change.
1
⌄
Observable events
What we could see on the surface
Sudden volume pressure
- Leadership mandated a full catalogue iteration
- Three-month window to redesign, create, test, and finalise
- Expected to redesign the entire existing library all at once
Quality & structural gaps
- Passive activities, lower-order thinking quizzes, low-utility worksheets
- A strict 60-minute structure introduced overnight: Theory → Case Study → Exercise → Q&A
- Upstream SME inputs lacked practical behavioural frameworks
Operational friction
- ID capacity still anchored to a thorough ADDIE workflow built for a different product
- Facilitators consistently received materials too close to live sessions, damaging delivery confidence
2
⌄
Patterns to anticipate
Recurring cycles beneath the events
- Volume–quality trade-off cycle: Each time volume targets spiked, quality shortcuts accumulated — creating a rework backlog that ultimately undermined the original speed gains.
- Awareness-level content ceiling: SME outputs consistently stopped at awareness (defining what learners should do), failing to scaffold the how.
- Mandate-driven development: Course direction was shaped by top-down mandates; user feedback and performance data had lesser influence.
- Rhythm collision: Deadline pressure repeatedly collided with the team's established ownership culture without a formal escalation path.
3
⌄
Design for change
What needed to be restructured
- Actionable scaffolding over dense theory — embed behavioural science directly into design templates; build case studies and exercises that push learners to apply skills immediately
- Lean workflow architecture — rebuild the sprint cycle to a 5–6 day maximum; create modular QA guidelines that explain the reasoning behind every standard
- Strategic stakeholder reframing — renegotiate the upstream SME mandate; frame operational changes around learner value and product risk to secure buy-in
- Working memory protection — ID leads consolidate and prioritise mid-sprint feedback before it reaches designers
4
⌄
Desired transformation
The system we were building toward
- From awareness to application — shift the product value proposition from passive consumption to practical exercises enabling learners to apply soft skills immediately, backed by post-course tools for self-directed growth
- From process mismatch to agile creation — transition from a heavy waterfall ADDIE review process to an optimised, high-velocity production sprint designed for short-form live content
- From execution silos to system alignment — align leadership expectations, SME outputs, and ID execution so the entire chain supports performance, not just output volume
- From reactive burnout to sustainable ownership — protect the team's psychological time and energy, modelling the human-centric design principles we deliver to learners
The Intervention
The redesign came down to shifting the organisational mental model from courses for awareness to courses for application — inside a live, high-volume production system. Click each step to expand.
1
⌄
Align the upstream inputs
Before any design work reached the team, I established a shared understanding with QA, SMEs, and senior leaders. The hardest part was negotiating changes to the SME report structure, as the shift was operationally costly. With the ID leads, I reframed the conversation around product risk and learner value — showing what would be lost if we maintained an awareness-only model.
2
⌄
Set the standards
Rather than asking designers to memorise a new approach while working at extreme speed, I built it directly into their daily templates. Slides were restructured for higher-order thinking. Case studies were scaffolded using behavioural science principles. Exercises shifted from memory recall to direct implementation. The team's feedback was essential — the template was actively designed and redesigned throughout.
3
⌄
Develop the capabilities
Team training modelled the exact design principles we wanted them to teach — focusing on the reasoning behind decisions, with hands-on practice paired with rapid QA feedback. The team was empowered to use AI tools (ChatGPT, Perplexity, Gemini) in course creation while owning the final output. The goal was true skill transfer, not strict compliance: if the team understood why a standard existed, they could adapt when facing unexpected design challenges.
4
⌄
Redesign the workflow
With my co-lead, I rebuilt the sprint workflow to survive the three-month timeline: a tight 5–6 day maximum cycle per course including QA and proofreading, progress tracked via Asana, and QA guidelines rewritten to show the logic behind every standard. Sprint assignments were dynamically adapted to team strengths in real time.
5
⌄
Manage cognitive load internally
When in-house trial feedback arrived during active production, passing it mid-sprint would have caused immediate cognitive overload. ID leads acted as an operational filter — consolidating raw feedback into actionable tasks during final reviews. I applied the same human-centric principles to our internal team that we designed for learners: protecting working memory during critical windows.
The Results
100%
Adoption of mandatory course structure across all redesigned courses
100%
Courses incorporating behavioural frameworks and practical application tools
100%
Courses meeting higher-order thinking standards
95%
Weekly delivery targets met within the 3-month production window
100+
In-house trial sessions establishing data-informed iteration cycles
5%
Gap intentional — adjusted for more specific learner segments
The 5% delivery gap was a deliberate choice: content adjustments were made to target more specific learner segments, prioritising accuracy over arbitrary speed.
Lessons Learned — Flashcards
Click any card to flip it. Switch themes using the filters below.
Tap to flip
What slows you down in a sprint — data gathering or poor prioritisation?
Poor prioritisation. Data gathering doesn't slow you down. The key is clear communication about what signals matter right now — revision queue, feedback lag, scope stability.
Tap to flip
Are speed and rigour opposites in high-velocity production?
No. They coexist when internal and external communication is clear about what data matters right now. In a 3-month sprint, measure only the signals that reveal system health.
Tap to flip
Is cross-team collaboration optional or structural?
Structural. Map each team's pain points, constraints, and mandates early. Misalignment doesn't resolve itself — it cascades across the entire production chain.
Tap to flip
What three signals reveal whether a production system is working?
Revision queue depth, feedback lag, and scope stability. These expose system-level problems before they become delivery failures.
Tap to flip
What bridges leadership vision and actual impact?
Systematic feedback. Build mechanisms to gather input from learners, facilitators, and SMEs that validate direction — or surface adjustments needed.
Tap to flip
What should be automated so humans can focus on judgement?
Admin reporting, database updates, feedback sentiment analysis, and pattern detection — freeing your team to synthesise data and make critical decisions, not process it.
Tap to flip
Why is tracking team wellbeing as important as tracking delivery?
Wellbeing is a performance variable. Cognitive load → decision quality. Attention fragmentation → rework cycles. Burnout → turnover and knowledge loss.
Tap to flip
How did feedback loops shift project direction?
100+ in-house trial sessions moved the iteration cycle from leadership-driven direction to data-informed cycles — facilitator and learner feedback became the primary adjustment signal.
Tap to flip
What are the four categories of change to track in any transformation?
People (roles, skill gaps, capacity), Process (workflow changes, handoffs), Tools (software, templates, automation), and Culture (mindset shifts, ownership, decision autonomy).
Tap to flip
How do you know if a process change is working if people aren't bought in?
Culture metrics will show it before delivery metrics do — feedback update rates, decision autonomy, and escalation frequency flag the gap early.
Tap to flip
Why is risk and impact assessment non-negotiable?
Map dependencies ruthlessly: which teams feed into your process, which processes feed into other teams' success, what external decisions affect your timeline. Gaps cascade.
Tap to flip
What's the right metric for tool adoption?
Time-to-productivity and bug reports — not just whether the tool is installed. These tell you whether the problem is training design or tool design itself.
Change Management Tracker
Four categories to track across this large-scale transformation. Filter by category.
| Category | Examples from this project | What to track |
|---|---|---|
| People | New ID roles, SME capacity constraints, skill gaps in short-form live design | Readiness, training completion, retention |
| Process | 5–6 day sprint cycles, SME report structure, facilitator handoff timing, QA guidelines | Cycle time, bottleneck movement, decision lag |
| Tools | Asana tracking, AI tools (ChatGPT, Perplexity, Gemini), redesigned design templates | Adoption, time-to-productivity, bug reports |
| Culture | Waterfall → agile mindset, data-informed iteration, ownership during transformation | Feedback update rate, decision autonomy, escalation frequency |