All work

Curriculum · Teaching · Systems

A five-month design curriculum

Built for people with no background, aimed at hireable

Over 1,000 designers trained across 15+ cohorts.

Measured: Instrumented. Number, source and date range shown.

Role
Curriculum Designer and Lead Facilitator
Timeline
2022 to 2024
Team
Multiple facilitators, cohorts of 60 to 200
Platforms
In-person and remote cohorts
Status
Running

Impact

  • Over 1,000 designers trained across 15+ cohorts

    Measured. Programme records, Tech4Dev

  • Around 60% of mentored students moved into design roles or freelance work

    Directional

  • Engagement and retention improved against the previous curriculum

    Directional

The problem

Most design education produces people who can follow a tutorial and rebuild a Dribbble shot, then freeze at a problem they haven't seen before. Students finish knowing Figma without knowing how to frame a problem, make a call under constraints, or defend it to a product team. Real projects start too late, feedback is thin, and the portfolios that come out are full of hypothetical redesigns that tell a hiring manager nothing.

What made it hard

  • Learners arrived with no design background at all, and had to leave employable.
  • Five months, fixed. Every week had to justify itself against being cut.
  • The bar was hireable, not finished. Completing the work wasn't the same as meeting the standard.
  • It had to be deliverable by other facilitators. If it only worked when I taught it, it didn't work.
  • Cohorts ran into the hundreds, so anything requiring one-to-one attention couldn't scale.

The decisions that shaped it

With what I turned down, and what each one cost.

01Real briefs from week two

Chose
Students work on live, constrained problems almost immediately, before they're comfortable.
Rejected
The conventional sequence: fundamentals first, projects once the tools are learned.
Why
Tools learned without a problem attached don't stick, and confidence built on tutorials collapses the first time a brief is ambiguous. Starting with a real problem is uncomfortable for about three weeks and then it compounds.
What it cost
Early attrition rose. Some people want to feel competent before being tested, and this curriculum doesn't offer that.

02Assessed on the defence, not the artefact

Chose
Students present decisions and get questioned on them. The reasoning is what's marked.
Rejected
Grading the final screens, which is faster, more consistent and far easier to moderate across facilitators.
Why
The screens are the easiest part to fake and the least predictive of whether someone can do the job. What a hiring manager is actually testing is whether you knew why. So that's what the programme tests.
What it cost
Much slower to assess and harder to keep consistent between facilitators. It needed a written rubric to stop it drifting into taste.

03Written down so someone else could run it

Chose
Full facilitator documentation: session plans, rubrics, worked examples of good and bad answers.
Rejected
Keeping the programme as slides plus my own delivery, which is how it started and was working fine.
Why
A curriculum that only one person can teach isn't a curriculum, it's a performance. Documenting it was the difference between a few hundred people and a few thousand.
What it cost
Several weeks of writing with nothing visible to show for it, at a point when the programme was already succeeding.

Delivered

What shipped. No outcome claimed.

  • Curriculum delivered by facilitators other than its author

What I'd do differently

I underestimated how much of the outcome came from assessment design rather than teaching. Changing what gets marked changed what students did far more than any lesson I rewrote. If I ran it again I'd design the rubric first and work backwards from there.

What the labels mean

Measured:
Instrumented. Number, source and date range shown.
Directional:
Observed in practice, not instrumented.
Target:
The goal the design was built against. Not yet measured live.
Delivered:
What shipped. No outcome claimed.