Dressing for work, without a dress code

A register is a band of formality, not a specific outfit — a range of garments a room reads as roughly equivalent, which is why two people at the same meeting can be dressed differently and both be correct. What matters is where the edges are: the least formal garment that still counts as business professional, the most formal that still counts as casual.

The table below is written by garment category rather than as two gendered lists, because the category is what a room reads. A structured shoulder over a clean top with a proper shoe reads as business professional whether the bottom half is trousers or a skirt. Splitting the advice in two doubles the length and halves its usefulness.

Read a row across and you have a complete outfit at that register. Mix rows and you land between the two, which is usually fine and occasionally the point: a suit jacket over a knit sits deliberately between business professional and business casual.

Since 2019 the whole scale has drifted about one full step downward, and fastest in the middle. What most offices now call business casual is what those same offices would have called smart casual before, and what they call smart casual is close to what used to be plain casual.

The written dress code is usually older than the practice. A handbook that says business casual was probably written when the phrase meant a collared shirt and pressed wool trousers, and has not been revised since. Dress to the document and you will be a step more formal than the room — a survivable error, and a smaller one than the reverse, but worth knowing you are making it.

Two registers held their ground. Business formal has not moved at all, because it is defined by specific garments rather than by a feeling, and a dinner jacket cannot drift. Business professional held where a jacket still signals something; the tie quietly went, so what was a suit and tie is now a suit, unchanged.

The gaps between registers are easier to see than to read about. These three are our own sample photographs, made with PoseLab, and each sits cleanly in a different band — the distances the table above describes, in garments rather than in words.

The single most useful heuristic here: when you do not know the register, dress one level above what you expect and no further. One level up reads as having taken the occasion seriously. Two levels up reads as having misunderstood it, and in some rooms as not knowing the people you are meeting.

It works because the cost is asymmetric. Overdressed is correctable in seconds: take the jacket off, undo the top button, push the sleeves back. Underdressed is not correctable at all, because you cannot produce a garment you did not bring. Make the mistake you can fix in the corridor.

Which is why the removable layer is the most valuable thing in a work wardrobe. A jacket you can take off is a register you can leave; a knit you can push up the forearm is a smaller version of the same trick. Buy for uncertain rooms accordingly.

You are rarely choosing a register in the abstract; you are dressing for one of a handful of recurring situations, each of which shifts the calculation. Below is the one-level-up heuristic applied to each, with the reason attached so you can adapt.

Changing industry is where the heuristic needs help, because you no longer have a reliable sense of what one level up even is. The usual failure is carrying your old register across intact: the lawyer moving into technology who wears the suit and is the only person in the building wearing one.

The fix is to observe rather than guess. Find the people roughly one level above you in the new field — not the founders, who are often exempt from the code, and not the most junior, who are still working it out — and look at how they present themselves publicly. Conference photographs, team pages and recorded talks are easier to read than a job description.

Then dress to the middle of what you see, not the top of it. The people at the top of any observed range are usually doing something specific: a photograph from a funding announcement, or a day on stage. Copying the outlier is how you end up two levels up rather than one.

The previous two chapters decide which garments to wear. This one is about whether they fit, and fit beats all of it. A cheap jacket that fits looks considered. An expensive jacket that does not looks like a jacket belonging to somebody else, and no amount of cloth quality, cut or provenance rescues it. People who cannot name a single thing that is wrong will still see that something is.

Clothing communicates through its lines. A shoulder seam in the right place produces a clean horizontal that reads as structure; the same seam an inch out produces a soft break that reads as slump. Nobody consciously registers the seam. Everybody registers the slump. That is why fit is worth more of your budget than anything else in this guide.

Four points decide almost all of it. Learn them properly: once you can see them you can assess a garment in about ten seconds, in a shop, before you have thought about colour or cloth or price — and three of the four can be corrected for a small fraction of what the garment cost.

Check these standing square to a mirror, arms hanging naturally at your sides. Not posed, not holding your breath — the posture you actually stand in, because that is the posture the garment has to work in. Most fit mistakes are made by someone unconsciously standing better than they normally do.

Three of the four points above are cheap to correct, and a tailor who does nothing else will improve a wardrobe more than a shopping trip will. Sleeve shortening, hemming and taking in a waist are routine work at any alterations shop, and cost a fraction of the garments they fix. So shop differently: buy to the point that cannot be altered, and have the rest brought to you.

For a jacket that means buying to the shoulder and having everything else adjusted. For trousers it means buying to the seat and thigh, which are structural, and having the waist and hem set afterwards. For a shirt it means buying to the neck and shoulder, then having the body taken in if it billows — a shirt that balloons under a jacket is common and entirely fixable.

What is not worth it: recutting a shoulder, moving a lapel, changing a jacket's overall length, or altering anything knitted. These are either impossible or cost more than the garment is worth. If a garment is wrong at the shoulder, it is wrong, and continuing to own it is a decision to keep being disappointed by it.

Sizing charts describe an average that very few people are. Broad shoulders with a narrow waist, a long torso, a short rise, one shoulder lower than the other — all ordinary, and none accommodated by off-the-peg sizing, which varies a single number and scales everything with it. If nothing ever fits, the likeliest explanation is not your body but the assumption built into the size run.

Buy to the widest or most structural measurement and have the rest brought in. That produces a garment too big in several places and correctable in all of them — a better starting position than one that is right in one place and impossible everywhere else. The label size is not information about you; it is information about that manufacturer.

The one asymmetry worth naming: a garment can nearly always be made smaller and rarely made larger, because letting out depends on seam allowance the maker may not have left. When you are between sizes and genuinely cannot decide, the larger one is the one with options.

"Avoid patterns" is one real phenomenon over-generalised. There are two ways a pattern fails under close scrutiny, they are unrelated to each other, and they point in opposite directions — which is why a single blanket rule cannot cover both.

The first is scale. A fine, regular repeat — a pinstripe, a hairline check, a tight birdseye — interferes with whatever grid is sampling it: a camera sensor, a screen, or the eye's limit across a room. What comes back is a crawling, iridescent flicker that reads as damage rather than as cloth. It depends on the ratio between the repeat and the size it is seen at: the same jacket is immaculate in person and broken in a small photograph.

The second is attention, and has nothing to do with resolution. A bold, irregular pattern does not shimmer — an irregular repeat has nothing regular to interfere with — but it becomes the second thing worth looking at in a situation that should have one. Hence the conclusion that surprises people: a tonal paisley is often safer than a high-contrast pinstripe. Scale and contrast decide, not the pattern's name.

The full treatment, band by band from solids up to architectural checks, is in our article on patterns in professional photographs. It carries a slider that shrinks a strip of swatches until the fine repeats break up and the irregular ones don't — the scale argument made better than any paragraph can.

Weight is the property nobody shops for and everybody notices. A heavier cloth hangs straighter, holds a crease, and returns to shape after a day of sitting down; a light one clings, creases where it folded, and shows every contour underneath. That is most of the difference between two visually identical trousers at six in the evening.

Drape is the separate question of how cloth falls when nothing is holding it. Good drape moves as a mass and hangs in long, soft lines; poor drape either stands away stiffly, holding a shape that is not yours, or collapses into small creases the moment you move. Test it in seconds: gather a handful of the fabric and release it. Good drape falls out and settles; poor drape keeps the shape of your fist.

So mid-weight beats light almost every time. Light cloth is a summer decision, not a default, and a cheap suit reads as cheap more because of weight than cut. If two garments look identical, pick both up: the heavier one will still look right in the afternoon.

Synthetic fibres are not the problem; blends solve real ones — they hold a crease, resist wear, cost less. Shine is the problem. Smooth synthetic surfaces reflect light in one direction rather than scattering it, so under a bright ceiling light they throw hard patches that a broken natural surface does not, and the garment stops reading as cloth.

The test: hold the garment under the strongest light available and turn it. If a bright highlight slides across the surface, it will do that in every bright room you take it into. Matte scatters light instead, and matte is the property to shop for — it matters more than the fibre content on the label.

The most underused option in work clothing is texture. A cable knit, a waffle, a twill line, an oxford basketweave, a linen slub, a corduroy wale, a flannel nap — each has a repeat too small to read as a pattern, so none of them carry the shimmer risk or the attention cost. What they give instead is depth: a textured mid-grey knit is more interesting than a flat one at exactly the same risk, which is none.

"Which colours suit you" is nearly unanswerable, and is where most of the confident nonsense in this field lives. "How much does this garment separate from what is next to it" is answerable, and it decides whether an outfit works at a distance, in a photograph, or across a table.

Three things are in play and two gaps between them: your skin and hair, the garment, and what is behind you. Each gap needs enough difference that the eye reads them as separate. Dark clothing against a dark wall loses the outline of your shoulders, leaving a head with no body; pale clothing under strong light against a pale wall fails from the other end. Neither is a colour being wrong; both are a gap being too small.

Mid-tones are the safe default because they sit between skin and most surroundings, so both gaps stay open without effort — slate, forest, warm grey, burgundy, olive. Not that they suit everyone — that they fail less often, being unlikely to collide with your skin or the wall behind you. The two conditions under which black and white are exactly right are in our piece on choosing colour for a headshot.

Colour is among the first things to go as viewing distance increases. Across a room, or in a photograph seen small, what survives is brightness and shape — your top being darker than the wall, the vertical line down each side from a jacket, the collar framing your face. Hue stops carrying information long before shape does.

Up close it inverts. At a meeting table, hue and finish become visible again and the failure modes change: a black that has faded towards brown next to a true black, two navies that are obviously two different navies — invisible from across a room.

So know which distance you are dressing for. A conference stage and a client meeting have different answers. If you have to satisfy both, dress for the far case, because the near case forgives more.

The seasonal palette industry — you are a Winter or an Autumn, so a fixed set of shades is right for you and the rest wrong — is presented as a method and works as merchandising. Different practitioners assign the same person to different seasons, and the classification does not survive changing your hair colour, which ought to end it as a theory about a person.

The real observation underneath is worth keeping. Some colours sit closer to your own skin and hair in brightness, and those give the eye less to separate. A garment very close in tone to your skin makes face and top read as one continuous area — which is what washed out means: a gap problem, not a season.

So keep the mechanism and drop the taxonomy. Hold a garment next to your face in daylight and ask one question: is there a clear difference in brightness? If yes, it will work in almost any room. If no, it needs help from a background or a layer — a decision you now make deliberately rather than inherit from a chart.

A working wardrobe does not need a set of clothes per register. It needs a few garments that each straddle two adjacent registers, so any pair of them lands somewhere useful. That is what makes a capsule small: the jacket that finishes a business casual outfit is the one that softens a business professional one, and the knit that dresses down a suit dresses up chinos.

Those garments are plain, mid-toned, matte and structurally simple, because anything with a strong identity of its own can only be worn where that identity is wanted. A patterned jacket is one outfit; a plain one is six. Not everything you own has to be boring — only the pieces doing structural work. Character comes from texture and from what you add later.

One warning: it is easy to buy a whole capsule at once and end up with clothes that fit a theory rather than your actual week. Look at what you did last month and build for that. If you have not worn a suit in two years, the jacket below can be an unstructured one and the list gets simpler.

Everything below is specified by category, colour and finish rather than by brand, because the brand is the least informative thing about a garment. Fit and cloth weight are what matter, and those you check in person against the four points in chapter three.

Once the list above is covered, add against gaps you actually hit rather than a sense that the wardrobe is small. The usual first additions are a second pair of trousers in a different weight, so summer and winter both work, and a coat that goes over the jacket without crushing the shoulder — an overcoat ending below the jacket hem.

Accessories are where a small wardrobe gains character or loses coherence, and the deciding property is scale. Small and matte disappears into an outfit; large and shiny becomes the thing people look at. A watch, a plain belt matched roughly to the shoes, and one piece of jewellery you actually like is a complete answer; the failure is quantity.

What never earns a place: anything with legible text, a visible logo or a novelty print, because people read text involuntarily. Anything shiny enough to catch a ceiling light. Anything that only works with one other garment you own, which is an outfit pretending to be a piece. And anything you keep because of what it cost — a sunk cost wearing a hanger.

Everything in this guide holds in front of a camera, only less forgivingly. A photograph freezes one moment, flattens depth, often crops at the shoulders and is usually seen small. Each of those narrows the margin: the fine stripe that shimmers slightly across a room shimmers unmistakably in a picture, and the pale top that was merely bright becomes the brightest thing in the frame.

The three variables carry over intact. Register still has to match the industry, and a suit in a field that stopped wearing them reads as a costume — the dress code by industry covers where the lines currently sit, including the regulated and uniformed professions most guides skip. Fit is more visible than in life, because a still image gives people time to look, and fabric is where chapter four's scale and attention problems do their damage.

Two things change enough to name. The crop means you are choosing about six inches of clothing, so the neckline and collar carry a weight they never carry in a room. And display size decides the pattern question outright, which is why what to wear for a LinkedIn photo reaches different conclusions from general headshot advice, and both are right.

PoseLab makes headshots from selfies you already have, and the wardrobe comes with the style you pick — so the choice moves from what you wear on the day to which register the finished photograph should sit in, which is the decision chapter one describes. How we choose the styles is an account of those wardrobe decisions rather than a pitch, including the ones we got wrong first. You can try it without paying: upload a few selfies and get one finished headshot back, no card.

What carries over into the selfies you upload is short. One flat colour, nothing across your face, nothing printed on the garment — chapter four's scale problem applies to what goes in as much as what comes out, so a fine check in a selfie works against you before anything else starts. The rest of this guide is about the room rather than the camera, and stands on its own whether or not you ever buy a photograph from anyone.