How to use this
The weekly rhythm
25 hours a week, split roughly:
| Block | Hours | What |
|---|---|---|
| Study | 6 | Reading, video, teardowns, note-taking against the module’s concepts |
| Make | 12 | The hands-on exercises — drawing, patterning, spreadsheet building, sample review |
| Reach out | 4 | Emails to factories, suppliers, material reps, potential mentors. This is the block people skip. Don’t. |
| Write | 3 | Your build log (below), plus the module deliverable |
The “reach out” block is non-negotiable and starts in Week 1, not Week 11. Supplier and factory relationships have a lead time measured in months. If you wait until you have a finished design to start emailing, you will lose eight weeks waiting for replies you could have already had.
Every module has the same skeleton
- Objective — what you can do afterwards that you couldn’t before
- Why it matters commercially — the money or risk consequence of not knowing this
- Core concepts — the actual content
- Do this — exercises with hour estimates
- Deliverable — a concrete artifact, added to your capstone folder
- Self-check — questions you should be able to answer cold
- Traps — the specific ways beginners get burned here
The build log
Keep a single running document from Week 1. Every week, write:
- What I learned that contradicted what I assumed
- Decisions I made and why (with the option I rejected)
- Open questions and who I need to ask
- Money spent this week, running total
This is not journaling. Three things depend on it: your tech pack accuracy, your ability to answer a factory’s “why did you spec it this way” question, and — later — your ability to tell the brand’s story truthfully, which is the only kind of brand story that survives contact with customers.
What to buy in Week 1
Starter kit, ~$180–350.
- 3 used sneakers to destroy, from a thrift store — deliberately different constructions: one vulcanized (Vans/Converse type), one cemented cupsole (skate or court shoe), one modern foam runner. $30–60.
- A hacksaw or Japanese pull saw, a sharp utility knife with spare blades, and a bench vice or clamp. $50–80.
- Digital calipers (0.01mm) and a flexible tape measure. $25–40.
- One pair of stock lasts in your sample size (typically US 9 men’s / EU 42) — search “shoe last HDPE” or ask a factory to send a pair. $30–90 including shipping.
- A dissection notebook (dot grid, A4) and a set of fineliners. $20.
- Optional but excellent: 1kg of scrap upper leather and a few sheets of 2mm and 4mm EVA foam. $40–70.
Everything else — software, patterns, materials — is introduced when the module needs it.
Software you’ll accumulate
| When | Tool | Cost | Why |
|---|---|---|---|
| W7 | Adobe Illustrator or Affinity Designer | $23/mo or $70 once | Industry-standard for footwear line drawings and tech packs |
| W7 | Procreate (iPad) or any sketching app | $13 once | Fast ideation on last profile templates |
| W9 | Google Sheets / Excel | free | BOM, costing sheet, cashflow — genuinely more important than the CAD |
| W13 | A shared drive (Drive/Dropbox) | free tier | Factories work in email + shared folders, not in Figma |
| W22 | Shopify | $39/mo | Only start paying when you’re within 6 weeks of selling |
| W1 | A general-purpose AI assistant (Claude, ChatGPT or similar) | free–$20/mo | Drafting, translation, spreadsheet building, first reads of classification and regulation. See Where AI helps below |
| W7 | A sketch-to-render tool (Sole Studio, NewArc or similar) | free tier, then credits | Colourway and material exploration before sampling. Not a tech pack |
You do not need 3D CAD (Rhino, Modo, Clo3D) or footwear pattern CAD (Shoemaster, Romans CAD, ICad3D+) for your first shoe. Factories pattern from a good 2D tech pack every day. Module 8 explains when 3D starts paying for itself.
Where AI helps, and where it lies
The tools above make a shoe. The ones in this section make you faster at everything around it. Used well, a language model and a render tool take hours out of every week of this plan. Used badly, they produce confident nonsense about shoes that don’t exist. One rule covers it: AI for drafting and exploring; physical samples, your broker and the factory for deciding.
| Phase | Where AI saves hours | Where it misleads |
|---|---|---|
| 1 · Fundamentals | Explaining a term or a construction you’ve just cut open. Quizzing you on the self-check questions. Turning dissection notes into your own glossary | It will describe a construction confidently and wrongly. Check it against the shoe in your hands, never the other way round |
| 2 · Design & product | Sketch-to-render tools turn a line drawing into colourway and material studies in minutes, so you try ten directions before paying for one proto. Drafting spec language and BOM structure for the tech pack. Translating it | A render is not a tech pack. Factories quote from 2D flats and a BOM. Renders also draw soles no open mold makes and constructions no small factory runs: the “wish, not a design” trap, now effortless |
| 3 · Sourcing & making | First-contact emails in Portuguese or Chinese. Screening a hundred listings for the tells in Module 12. Live translation on video calls. Writing sample comments that can’t be misread. Reading a quote breakdown | Invented factory names, addresses, MOQs and lead times. Every supplier fact comes from the supplier, confirmed on a video call. An agent’s translation value has fallen; their factory access and QC presence have not |
| 4 · Money & law | Building the seven-tab model in an afternoon. Scenario tables. A first read of the HTS and CN headings. A plain-language summary of REACH or GPSR. Drafting the one-page plan | Invented tariff lines and rates. Classification is your broker’s call, not the model’s. Legal summaries drift out of date; verify against the current text within eight weeks of shipping |
| 5 · Brand & market | Product copy and size-guide drafts. Background removal and clean cut-outs. Email sequences. Translating the product page for a second market | Generated images of a shoe that doesn’t exist yet. Pre-selling on them is a consumer-law exposure in both markets and a returns disaster. Photograph the real sample on real feet |
| 6 · Ship & learn | First drafts of customer-service replies. Tagging return reasons. The post-mortem’s numbers. Fit-prediction and size-recommendation tools on the store, which attack the largest cost in online footwear | A chatbot that promises what your policy doesn’t. Write the policy yourself; let the model quote it |
What it does not change
The last and the fit. Material minimums and factory MOQs. Ninety to a hundred and twenty days of production. Freight, duty and the cash cycle. The sample rounds in Module 14, every one of which is a physical object arriving in a box. Information work got cheaper. Pairs, molds and containers did not. The sub-$10k floor in this curriculum is set by tooling, minimums and freight, not by hours, and no tool moves it.
The one technology that does attack that floor is on-demand 3D printing. Platforms now print a shoe to a customer’s foot scan with no inventory, no tooling and no size run, and some accept outside designers. The product is printed polymer, not a leather cupsole, and it is a different aesthetic for a different customer. Treat it as a side experiment or a V2 idea, not as a substitute for the path this curriculum takes.
Three things not to do
- Don’t pre-sell on renders. Pre-orders open when a confirmation sample exists and has been photographed on feet (Module 19).
- Don’t let a model write a number into your spreadsheet that you can’t trace to a quote, a broker, a carrier or a test. The model is only useful if every input has a source.
- Don’t send a spec, a sample comment or an email you haven’t read as if you wrote it. The factory holds you to it, and “the AI wrote that” is not a defence anyone in a sample room will accept.
Where the saved hours go: back into the Make and Reach out blocks of the weekly rhythm, not into more time with the tools.
A note on honesty with yourself
Two failure modes kill first brands, and both are avoidable by being blunt in your build log:
- Designing something you cannot make. A design that requires custom tooling, exotic materials or a construction your factory has never run is not a design; it’s a wish. Constraint-first design is a skill and Module 7 treats it as one. Render tools make this failure effortless, which is why the section above exists.
- Confusing an audience with a market. People liking your Instagram is not evidence they will pay $140. Module 19 forces you to test the difference with money before you commit to a run.