Design Development

Thumbnail sketching

Thumbnail sketches are small, fast drawings used to make many design possibilities visible before time is invested in detail or rendering. Their value lies in comparison: scale, repetition and low commitment allow one idea to be tested beside another. Stanford d.school defines ideation as “going wide” to create both volume and variety, and recommends separating idea generation from evaluation so early judgement does not close the field too soon.^[1]

Begin with a written design question, not a blank page. State the wearer or participant, use situation, collection concept, garment category, material direction, key constraints and the quality to explore. A useful question is specific enough to focus action but open enough to permit difference: “How might a commuter coat create adaptable weather protection without a conventional fitted waist?” is more generative than “Design a black coat.”

Set up the page as a field of equal frames. Use one figure scale, one view and one time limit for the first batch. Number every sketch in the order made; add no shading or decorative background. Work with a tool that encourages speed—a soft pencil, marker, brush or simplified digital brush—and make the garment readable through outer contour, major internal division and one defining detail. A thumbnail is successful when another designer can understand its proposition, not when it looks finished.

Sketch in deliberate rounds:

1. Release the obvious: record immediate solutions quickly so they do not occupy the rest of the session.
2. Change one variable: alter length, width, closure, volume, balance, opening, layer or body relationship while holding the rest stable.
3. Force distance: invert a rule, remove a familiar component, change garment category, exaggerate scale or translate a research shape into construction.
4. Combine: join the strongest principles from two unrelated thumbnails rather than tracing either one.
5. Return with evidence: add material, movement or wearer findings and generate a new batch.

Use silhouette first. Fill the garment shape in black or draw only the contour so small differences remain visible. Then add a second pass for internal architecture—seams, openings, panels, fastenings and layers. Add details only when they change function, construction, proportion or concept. Pockets copied onto every option are not development; a pocket that repositions volume, enables transformation or creates a family code is.

Annotate decisions, not appearances. Beside each thumbnail, note the design action, intended material behaviour, construction question and next test: “side panel expands through pleat,” “bias grain must be verified,” “closure accessible with one hand,” or “volume may collapse in soft jersey.” Written language helps the sketch become a working instruction and prevents attractive ambiguity from being mistaken for resolution.

Alternate lateral and vertical development. Lateral sketching moves to a meaningfully different concept; vertical sketching deepens one concept through detail and feasibility. Research comparing novice and expert conceptual sketching examines this interaction, along with sketch quantity and complexity, as a central part of idea development.^[2] Mark each move on the page so you can see whether you are exploring the field or polishing one branch.

Protect the early stage from fixation. Delay exposure to highly resolved reference garments when possible; extract principles from research—rhythm, suspension, folding, protection, growth—rather than copying a visible solution. A controlled 2024 study found that participants using AI-generated imagery produced fewer ideas with less variety and originality than a no-support baseline in its visual ideation task, while also showing greater fixation on the supplied example.^[3] This does not make AI unusable; it means its outputs should be treated as hypotheses, deliberately countered and not allowed to become the first or only visual vocabulary.

Select only after generating enough contrast. Cluster thumbnails by underlying principle, not superficial resemblance. Identify repeated defaults, missing categories and outliers. Score the most promising options against concept, wearer value, silhouette, material, construction, movement, responsible use of resources and collection role. Advance several different directions, including one riskier option, rather than selecting five near-duplicates of the same favourite.

Translate selected thumbnails without cleaning away their energy. Enlarge them, correct balance and add front, side and back views. Test the essential proposition immediately in paper, fabric, on the stand or through a simple digital volume. FIT’s fashion-design curriculum connects rapid and accurate sketching with proportion, details, silhouettes, line planning and design groups, treating drawing as part of collection development rather than an isolated illustration exercise.^[4]

Keep the complete sequence. Date and number pages, link each refined drawing back to its source thumbnail and record why a branch stopped. Failed options demonstrate judgement when the reason is visible. A development portfolio should show that the outcome emerged from alternatives, tests and edits—not that the final image appeared fully formed.

Studio exercise: Write one design question and produce forty thumbnails in four ten-minute rounds. Each round must use a different trigger: one-variable change, inversion, research translation and combination. Cluster the results, select four contrasting options and prototype the defining principle of each with paper or scrap fabric.

Portfolio checkpoint: Submit the design question, numbered thumbnail sheets, cluster map, fixation audit, shortlisting matrix, enlarged front/side/back studies and four rapid prototypes. Annotate what changed between rounds and why two promising ideas were not advanced.

Silhouette exploration

Silhouette is the perceived outer shape of a clothed body, including the spaces between the body, garment, limbs and environment. It is read before surface detail and often carries the strongest collection signal. FIT teaches garment silhouette through fashion sketching and develops both structured and soft silhouettes through draping, construction and material behaviour.^[4][5]

Analyse silhouette in more than a front-facing pose. Draw or photograph front, profile, back and three-quarter views, then include sitting, walking, reaching and layering. A side projection, back volume or gap under the arm may define the design more than the frontal outline. Check positive form and negative space: the opening between sleeve and torso, space inside a cut-out, stride between trouser legs and air beneath a suspended layer all contribute to the read.

Build a silhouette vocabulary from observable geometry rather than fixed body stereotypes. Useful starting structures include column, A-line, trapeze, cocoon, hourglass, wedge, sphere, rectangle, inverted triangle and asymmetrical mass. These are analytical shorthand, not rules or gender assignments. Combine, distort or localise them: a narrow column may carry one spherical sleeve, a cocoon may open into a sharp wedge, and an A-line may exist only in the back view.

Explore the variables systematically:

• Extent: close to body, skimming, loose, oversized or supported away from the body.
• Distribution: where volume begins, reaches maximum and returns—shoulder, chest, waist, hip, back, sleeve or hem.
• Direction: vertical fall, horizontal spread, diagonal pull, spiral, upward lift or downward drag.
• Edge: crisp, soft, broken, transparent, fringed, layered or moving.
• Support: body, fabric stiffness, gathering, padding, boning, pleating, air, fastening or internal structure.

Run black-shape tests. Fill each proposal as a solid silhouette at the same scale, remove colour and compare it at full-page and thumbnail size. Ask whether the concept remains legible, whether one area dominates unintentionally, and whether two designs are truly different or merely decorated differently. Repeat as white shapes on black to expose negative spaces and openings.

Let material participate early. Drape a stable paper, a fluid cloth and a resilient knit into the same nominal shape; photograph the differences. Fabric weight, bending resistance, stretch, recovery, thickness, friction and grain determine whether an outline stands, hangs, clings or moves. FIT’s draping courses explicitly study drapability as an influence on silhouette and style and use draping and flat pattern to solve design-development problems.^[6]

Move between 2D and 3D rather than treating the sketch as a command. Cut paper volumes, pin directly on a stand, build half-scale shapes, modify a block, scan the result back into a drawing and repeat. UAL’s research on historical fashion uses making, unmaking, remaking and digital animation to understand how pattern-cut forms become garments and move on the body.^[7] The translation itself can reveal the design.

Distinguish body transformation from body exclusion. Padding, compression, support, exposure and visual emphasis can be valid conceptual tools, but evaluate comfort, mobility, breathing, sitting, dressing and cultural meaning. Test on bodies or avatars representing the intended population rather than a single idealised croquis. The Museum at FIT documents how fashion has historically shaped an “ideal” body and marginalised other body types, while garment cut directs attention and creates desired silhouettes.^[8]

Research historical silhouettes through construction and context, not outline copying. Identify the understructure, material, body practice, class, labour, politics and technology that made a form possible. The Met’s study of nineteenth-century dress shows how petticoats, cage crinolines, bustles and corsetry redistributed volume and changed the apparent waist and body.^[9] Translate a principle with attribution and contemporary purpose rather than detaching a shape from its history.

Evaluate silhouette against the collection brief. Does it communicate at distance? Does it serve the wearer and function? Can it be produced in the chosen material and price structure? Will it work across sizes, layers and movement? Does it repeat a family code or provide necessary contrast? Keep a silhouette only when its visual force and practical consequences are both understood.

Studio exercise: Develop one research principle into twelve black silhouettes: three close-to-body, three suspended, three asymmetrical and three transformed through movement. Realise four in half scale using contrasting materials, then redraw what the material actually did rather than what the first sketch predicted.

Portfolio checkpoint: Submit front/profile/back silhouette sheets, negative-space tests, geometry and volume map, material comparisons, half-scale photographs, movement sequence, body-and-access review and a final silhouette rationale linked to the collection.

Proportion studies

Proportion is the relationship of one dimension, area or mass to another. In fashion it includes garment to body, one garment section to another, detail to whole, one layer to another and one look to the complete collection. Proportion is not a formula for an “ideal” figure; it is a design system for controlling emphasis, balance, rhythm and function on the intended range of bodies.

Establish measurable anchor points: crown, shoulder, bust or chest, underarm, waist, high hip, full hip, crotch, knee, ankle, wrist and floor. Add garment-specific points such as collar depth, pocket opening, break point, rise, sleeve fullness, hem circumference and layer overlap. Use a consistent reference height in drawings, but label which lines are body landmarks and which are garment decisions.

Separate three kinds of study:

• Vertical proportion: lengths and intervals—torso to leg, jacket to skirt, rise to inseam, yoke to body, sleeve to hand, hem to floor.
• Horizontal proportion: shoulder, chest, waist, hip, sleeve and hem widths; ease and the relationship between fitted and expanded zones.
• Surface proportion: the relative visual area of colour, print, transparency, texture, fastening, pocket and negative space.

Make a proportion ladder. Draw the same base design at equal scale, then move only one line through a planned sequence—for example, five hem lengths, five waist placements or five pocket scales. Mark the ratio or measurement under each version. This isolates cause and effect: when several variables change at once, it becomes impossible to know why the balance improved.

Use optical principles deliberately. A strong horizontal division can shorten or widen a zone; uninterrupted vertical movement can lengthen it; a concentrated dark or dense area may carry more visual weight; repeated divisions create rhythm; and scale changes can make a detail appear structural or trivial. Treat these as contextual tendencies, not universal laws. Material, colour contrast, movement, body shape and viewing distance can reverse the expected effect.

Ground studies in anthropometric evidence. ISO 8559-1 describes body measurements used to create anthropometric databases, select market populations and develop size-and-shape profiles and fit mannequins.^[10] ISO 8559-3 explains that body proportions differ significantly within targeted populations and provides a methodology for measurement tables and intervals.^[11] A sample-size mannequin is one testing tool, not a complete representation of the customer.

Test proportion across at least three relevant sizes or body shapes before freezing the design. Scale, grade or reposition elements according to their function rather than enlarging every detail uniformly. A pocket must remain usable; a collar must relate to the neck and face; a print or motif must retain its intended visual role; an opening must allow dressing. Document whether each feature is fixed, graded, repositioned or offered in variants.

Compare nude-body, underlayer and dressed-garment measurements. Ease, padding, compression, underwear, knit recovery and layering change the apparent relationship. Photograph the prototype on a neutral background from a consistent camera height and distance; a wide-angle lens or elevated viewpoint can distort conclusions. Check the design in a mirror, at distance and in motion as well as through close-up images.

Move beyond the front view. A long back, side drape, forward shoulder, dropped crotch or projected pocket changes the total balance even if the front appears conventional. Draw sectional views and volume maps, then test sitting, reaching and walking. Proportion must support movement, access to pockets and closures, modesty or exposure decisions, and the intended relationship between wearer and space.

Use the collection line-up as a final scale test. Arrange looks at equal size and inspect hem rhythm, outerwear length, waist placement, volume distribution and focal areas. Too much uniformity can make the range static; uncontrolled shifts can make it incoherent. Decide where proportion is a repeated code, where it evolves and where a deliberate break creates emphasis.

Studio exercise: Select one look and create three ladders: hem length, waist placement and focal-detail scale. Combine the strongest relationships into three proposals, then test them on three relevant body shapes or sizes. Record which proportions remain fixed and which require functional adjustment.

Portfolio checkpoint: Submit anchor-point diagrams, vertical/horizontal/surface studies, proportion ladders with ratios, multi-size or multi-body tests, front/side/back photographs, movement checks and a line-up rhythm analysis. State how evidence changed at least one initial preference.

Design variations

A design variation is a purposeful alternative generated from a clear concept, not a copy with random decorative changes. Variations help reveal which attributes are essential, which are flexible and which combination best serves the wearer, collection and production brief. The Design Council’s Double Diamond describes development as producing different answers, testing them, rejecting weak solutions and improving promising ones.^[12]

Define the design’s invariants before varying it. These may be a relationship to the body, construction logic, repeated curve, transformation mechanism, material contrast, opening sequence or emotional effect. Then list variables that can change: garment type, length, volume, collar, sleeve, closure, seam map, pocket, layer, colour, fabric, finish, print scale, styling and degree of exposure. If everything changes, there is no meaningful comparison; if nothing structural changes, there is no meaningful development.

Build a variation matrix. Place variables in rows and clearly different options in columns. Generate first within one row—five closure systems or five sleeve-to-body relationships—then combine selected options across rows. Mark incompatible combinations and explain why. The matrix is a prompt, not an automatic generator: every combination must still express the concept and function as clothing.

Use transformation operations to move beyond familiar styling:

• Extend or compress: change scale, duration, coverage or distance from the body.
• Shift: relocate a seam, opening, volume, fastening or focal point.
• Rotate or reverse: change orientation, face, sequence or front/back logic.
• Remove: delete a conventional part and solve the function another way.
• Combine or split: merge garment types, layers or details, or separate one component into a modular system.
• Translate: express the same principle through a different material, construction or category.

Work in divergent and convergent cycles. In a divergent cycle, aim for difference and suspend ranking. In a convergent cycle, group, compare, prototype and edit against written criteria. Return to divergence when tests expose a new question. This is development, not indecision: each cycle should reduce uncertainty or deepen the concept.

Vary across representation modes. A flat sketch can expose seam logic; a fashion figure shows body relationship; a black silhouette clarifies mass; collage tests scale and juxtaposition; draping reveals material behaviour; a paper model tests volume; a pattern sample reveals geometry; and a digital garment can compare colour or volume quickly. Parsons describes fashion development as an integrated 2D, 3D and 4D process, including visual communication, material research and systems thinking.^[13]

Prototype the highest-risk difference, not the entire garment each time. If variations differ by collar suspension, mock up the collar and shoulder. If the question is access, test the opening and fastening on the intended body. If volume depends on cloth, build the full unsupported area in the real or closely matched material. UAL’s Fashion Design and Development programme uses pattern cutting, construction, prototyping, sketchbooks, development files, toiles and 3D outcomes as connected evidence of creative product development.^[14]

Evaluate through a visible matrix. Criteria may include concept strength, wearer benefit, silhouette, proportion, movement, material compatibility, construction, responsible resource use, cost, sizing, collection role and communication clarity. Weight the criteria for the brief; a protective uniform and an editorial showpiece should not use identical priorities. Record evidence, uncertainty and the next action beside the score.

Use critique as information, not a vote. Ask observers to describe what they see, where attention goes, how they think the garment works and what remains unclear before revealing your intention. Separate preference from evidence. One comment can reveal a real failure; ten comments can still be irrelevant if the observers do not represent the intended wearer or context.

Maintain version control. Give each direction a family code and each iteration a sequential number. Photograph prototypes consistently, attach material and pattern references, date decisions and note what changed. When combining versions, identify the source of each feature. This prevents approved elements from disappearing and makes technical handover possible.

Know when to stop. A variation is ready to advance when its central proposition is legible, major risks have been tested, its collection role is distinct, and remaining questions can be solved through normal development rather than a new concept. Archive the rest. Unselected ideas can seed another product, but adding them all to the current design usually weakens it.

Studio exercise: Choose one thumbnail and define three invariants plus eight variables. Generate a thirty-option matrix, select six genuinely different combinations and prototype one high-risk feature from each. Use weighted criteria to advance two options and combine them only if the design logic remains clear.

Portfolio checkpoint: Submit the invariant statement, variation matrix, transformation studies, divergent/convergent map, six prototypes, critique notes, weighted evaluation, version history and final development sequence. Make every rejection traceable to evidence or the brief.

Garment families

A garment family is a group of products connected by recognisable design DNA, material or construction logic and a clear role within the collection. A family is more than several garments styled together, and it is more than one design repeated at different lengths. Each member should be identifiable on its own, related to the others and useful within the total range.

Write the family code in one sentence: “adjustable outer layers built from offset rectangles and one visible fastening,” or “soft base pieces defined by a diagonal wrap line and controlled gather.” List three to five visual or functional signatures, then state what must vary. This creates a test: if a new garment needs unrelated motifs to appear connected, the family logic is too weak.

Give each member a role:

• Foundation: accessible, versatile pieces that carry the collection language across frequent use.
• Bridge: products that connect colour, material, occasion or price groups.
• Statement: high-impact expressions that communicate the concept and attract attention.
• Layer: products that alter proportion, climate response or styling combinations.
• Completer: garments or accessories that make the wardrobe proposition usable.

Start with a family map, not individual illustrations. Place categories, intended wearer needs, climate, delivery, price tier, colour, fabric, silhouette, layer and styling relationships in a grid. UAL’s collection-building guidance joins purpose, target market, storyboarding, range size and diversity, material, colour and silhouette, then selects, alters and redraws designs to achieve a balanced collection.^[15]

Create unity through controlled recurrence. Repeat an edge shape, seam direction, pocket logic, fastening, proportion, fold, material contrast or method of transformation. Change its scale and function across categories: a diagonal closure might become the coat opening, dress wrap, trouser pocket and bag flap. Avoid attaching a logo-like detail after the fact; the strongest family codes participate in construction and use.

Create difference through category, coverage, volume, material, colour and price. Arrange family members as a progression rather than clones. One may be the purest expression, one the most wearable, one the most technically advanced and one the most experimental. Check that removing any member leaves a genuine gap; if not, it may be redundant.

Build technical commonality where it adds value. Shared fit blocks, sleeve systems, pocket bags, hardware, linings, seam types, materials and trims can improve consistency, reduce development and simplify repair or production. Commonality should not force unsuitable construction onto every category. Record platform elements that are fixed, those adapted by product and those unique to the statement piece.

Test outfit and wardrobe logic. Combine every top with more than one bottom, every layer with the intended underpieces and every closure or volume in motion. Photograph a combination matrix rather than only preferred runway looks. Check colour and material across front, back and reveal states. A family should support real choices while preserving the collection’s point of view.

Plan bodies and sizes as part of the family. Decide which proportions remain constant, which details grade, and whether lengths, coverage, closures or support require options. Test shared design DNA across the intended size and shape profile; do not reserve the statement language for one sample body while simplifying other sizes beyond recognition. Use clear body data and fit trials, then document exceptions by function.

Balance creativity with assortment. FIT’s fashion product-development curriculum connects colour, silhouette and pattern sourcing with assortment planning, while its fashion-design courses apply design development across categories, markets and price points.^[16][4] Count products by category, material, colour, price, delivery and role; identify overconcentration, missing entry points and manufacturing bottlenecks before the line-up is final.

Edit at three distances. From far away, judge silhouette rhythm and repetition. At page distance, judge category, colour, layer and material balance. Up close, judge construction signatures and detail hierarchy. Remove or redesign any member that adds quantity without a distinct wearer purpose or collection function. Strength comes from relationships, not maximum product count.

Document the family for handover. Use a range plan with style code, thumbnail, category, family, colourway, material, size range, price target, delivery, prototype status and key shared components. Link every style to its development file and technical drawing. Name the approved family code and version so later substitutions do not erode the collection logic.

Studio exercise: Develop one five-piece family from a single design principle. Include a foundation, bridge, statement, layer and completer. Create a family DNA sheet, range grid and outfit matrix, then remove one item and prove whether the family becomes weaker—or whether that item was redundant.

Portfolio checkpoint: Submit the family statement, signature/invariant list, five product developments, role map, shared-component platform, outfit matrix, multi-size checks, assortment audit, edited line-up and final range plan. Explain how coherence was achieved without repetition.