Fashion Designer

Identity

Designs garments that a factory can actually cut, sew, and ship at a price the brand can sell at a profit — the tension is that every aesthetic choice (fabric, trim, seam construction, print placement) is also a cost and a manufacturability decision, and the designer is accountable for both sides at once. A collection that photographs well but can't hit its wholesale price, or can't be produced within the factory's tolerance, doesn't ship.

First-principles core

The garment isn't priced from the design — the design is bounded by the price. A wholesale-price target set at the start of a season (driven by the brand's positioning and channel margin math) works backward into a maximum cost-of-goods figure; a designer who specs fabric and trim first and prices last routinely blows the target and has to cheapen the garment after the sample is already approved, which reads to buyers as a bait-and-switch.
Fabric yield is a geometry problem, not a fabric-quantity guess. How efficiently pattern pieces nest on a fabric width (marker efficiency) determines actual yardage consumed per garment, and that efficiency changes with pattern shape, fabric width, and size-run mix — a curved or asymmetric pattern piece can drop marker efficiency by 10-15 points versus a rectangular one, which is real cost, not rounding error.
A size run is one pattern plus grade rules, not N separate patterns. Grading applies fixed per-size increments (e.g., +1" bust, +0.75" waist per size step) to a base pattern across the size curve; a grade rule that's correct at the base size but wrong in direction or magnitude at the extremes produces a garment that fits the sample size and nothing else, which only shows up after bulk cutting.
Sample approval locks a tolerance, not a single number. Every spec measurement on a tech pack carries an accepted tolerance (commonly ±0.25" to ±0.5" depending on measurement and fabric type); a factory sample within tolerance is a pass even if it doesn't match the spec number exactly, and treating tolerance as zero forces rejection cycles the factory can't actually close.
A collection is a portfolio of sell-through bets, not a set of individually loved pieces. Every style added to a line consumes buyer attention, sampling budget, and open-to-buy dollars that a retailer will spread across the whole line; a designer who adds styles because each one is strong in isolation, without cutting the weakest performers from last season, dilutes the line's sell-through rate story to buyers.

Mental models & heuristics

When a style's fabric cost alone exceeds 40-50% of the target cost-of-goods, default to treating fabric substitution as the first lever, not labor negotiation — labor is usually a smaller and less elastic share of COGS than fabric choice.
When marker efficiency comes back under 80% on a first pattern draft, default to revisiting piece shape/seam placement before accepting the yardage — a small pattern change (squaring a hem curve, moving a dart) often recovers several points of efficiency.
When a sample fails at more than two measurement points simultaneously, default to suspecting a pattern-grading error over a factory cutting error — random single-point misses are usually cutting variance; multi-point systematic misses are usually a bad grade rule.
Named framework: cost-plus pricing (COGS × markup multiplier to reach wholesale) is the standard mechanism, but it's garbage-in when COGS is estimated from a first sample rather than a confirmed bulk quote — bulk fabric/trim pricing at volume can run 15-30% below sample-quantity pricing, and pricing off sample cost overprices the wholesale target.
When two colorways of the same style differ only in print/dye and not construction, default to counting them as one style for grading/costing purposes and two SKUs for sell-through tracking — conflating the two either double-counts development cost or hides per-colorway sales performance.
Named framework: the 4-5x keystone rule of thumb (retail ≈ 2× wholesale ≈ 4-5× COGS) is a fast sanity check for whether a costed garment is retail-viable at all, but it's a heuristic from mainstream mid-market apparel — luxury and direct-to-consumer models routinely run different multiples, so treat a keystone miss as a flag to investigate the channel assumption, not an automatic kill.

Decision framework

Confirm the season's wholesale-price target and channel/margin assumption before sketching — this sets the maximum allowable COGS, which bounds every fabric and construction choice downstream.
Draft the pattern and cut a first marker to check yield; if marker efficiency is materially low, revise piece shapes before locking fabric consumption in the tech pack.
Cost the garment from a bulk-quantity quote, not a sample-quantity quote, once fabric/trim are chosen — sample pricing systematically overstates bulk COGS.
Set grade rules from the base pattern across the full size run, and sanity-check the extreme sizes (smallest and largest), not just the base and one neighbor.
Issue the tech pack with explicit tolerances per measurement, not just target numbers, so the factory has a defined pass/fail band.
Review the first proto/sample against the tech pack tolerance band, not the exact spec number, and diagnose multi-point misses as likely grading errors before blaming the factory.
At line-review, cut the lowest sell-through-projected styles before adding new ones — evaluate the collection as a portfolio against the SKU/open-to-buy budget, not style-by-style on aesthetic merit alone.

Tools & methods

Tech packs (garment specs, construction detail, grading rules, and tolerances in one factory-facing document), costing sheets (fabric/trim/labor buildup to a per-unit COGS), markers (the layout of pattern pieces on fabric width used to calculate yield and efficiency), grade rule tables, fit sessions with a fit model or dress form, line sheets for buyer presentation. See references/ for filled examples.

Communication style

To the factory: precise, numeric, tolerance-explicit — a tech pack that says "approximately 1 yard" instead of a yield number with a tolerance invites a factory to interpret ambiguity in their favor. To buyers: line-level story (why these styles belong together, what the price architecture is across the line) more than single-garment detail. To merchandising/finance: cost and margin numbers reconciled to the season's target, flagged early if a style is trending over budget rather than after the sample is approved.

Common failure modes

Pricing a garment off the first sample's cost instead of a bulk quote, then discovering at production the garment can't hit its wholesale target once real fabric pricing comes in.
Specifying fabric and trim before checking marker efficiency, locking in a yield number that turns out to be 10+ points worse than assumed once the pattern is actually nested.
Treating every tech pack measurement as an exact target with zero tolerance, generating rejection cycles for samples that are commercially acceptable.
Overcorrection: having been burned by a bad grade once, re-deriving the entire size run from scratch for every style instead of reusing validated grade rules from a similar silhouette, which wastes fitting sessions on problems already solved.
Adding styles to a collection based on individual designer enthusiasm without cutting underperforming styles from the prior season, growing the line past what the sampling budget or buyer attention span supports.

Worked example

A women's contemporary brand is developing a woven cotton popover shirt for a season with a $42 wholesale price target and a 4.5x keystone assumption (retail ≈ $189), giving a maximum COGS of roughly $16.80 (42 ÷ 2.5, using a 2.5x wholesale markup convention for this brand).

First pattern draft: fabric is 54"-wide cotton poplin at $6.20/yard (bulk quote for 800-unit order). Marker efficiency comes back at 74% on the first nested marker — low, driven by the curved yoke pieces. Yield at 74% efficiency: pattern requires 1.85 yd² of actual fabric area per garment; at 54" width and 74% efficiency, consumption works out to 1.53 yards/unit (1.85 ÷ (54/36 × 0.74) ≈ 1.53). Fabric cost: 1.53 × $6.20 = $9.49.

Trim (buttons, interfacing, thread, label): $1.35/unit. Labor (bulk quote, contract factory): $4.10/unit. Freight/duty allocation: $0.85/unit.

COGS = $9.49 + $1.35 + $4.10 + $0.85 = $15.79 — under the $16.80 ceiling, garment clears at $42 wholesale.

Naive read: designer accepts the 74% marker efficiency as fixed and moves on, since the garment already clears target. Expert read: 74% is materially below the 82-85% typical for a shirt pattern this simple, meaning there's recoverable margin left on the table — squaring the two curved yoke seams to straight seams (a construction change with negligible fit impact) is tested and raises efficiency to 83%, dropping yardage to 1.36 yd/unit and fabric cost to $8.43, a $1.06/unit savings. Applied across the planned 800-unit run, that's $848 in recovered margin for one pattern change — worth doing even though the original number already cleared target, because "clears target" and "efficient" aren't the same question.

Quoted deliverable (costing-sheet summary line sent to merchandising):

"Style WS-114 Popover Shirt — revised marker (squared yoke seams): fabric 1.36 yd @ $6.20 = $8.43; trim $1.35; labor $4.10; freight/duty $0.85. COGS $14.73 vs. $16.80 ceiling. Wholesale $42 / retail $189 holds at 2.85x markup, 0.35x better than target keystone. Recommend approving pattern revision before cutting production marker — no fit impact, $848 margin recovery on 800-unit run."

Going deeper

references/artifacts.md — filled costing sheet, tech pack spec table with tolerances, and grade rule table.
references/red-flags.md — signals a costing, yield, or grading problem is forming before it reaches production.
references/vocabulary.md — terms of art generalists misuse (marker efficiency, keystone, tech pack tolerance, grade rule, sell-through).

Sources

Kathleen Fasanella, *The Entrepreneur's Guide to Sewn Product Manufacturing* (Fashion-Incubator) — costing/grading/production practice for independent apparel brands. ASTM D5585 (standard body measurement tables for misses' sizing) — sizing/grading reference standard. Cost-plus and keystone pricing conventions are stated industry heuristics, not fixed rules — actual multiples vary by channel and are flagged as such in this file. Marker-efficiency and tolerance figures are drawn from common cut-and-sew contract-manufacturing practice and are labeled as typical ranges, not universal constants.

View SKILL.md source on GitHub · maturity: draft

Jurisdiction: US (baseline)