What to wear to a job interview in 2026, on camera and in person

Wear one clean step above what the team wears on an ordinary Tuesday, and let the format decide the rest. That is the entire rule, and most of the advice you will read is a longer way of saying it. What follows is how to work out what that Tuesday actually looks like at the place you applied to, what genuinely changes when the interview arrives through a webcam, and which of the two you are likely to get.

Will your interview be on camera or in a room?

Nobody publishes a count of job interviews by format. The federal statistics that come closest measure where people work, and that turns out to be a decent proxy, because employers who run a distributed team run distributed interviews.

In August 2026, the Bureau of Labor Statistics counted 21.6 percent of people at work as having teleworked or worked at home for pay. Split that by occupation and the average dissolves. Computer and mathematical occupations came in at 60.8 percent; office and administrative support at 22.8 percent; sales and related at 23.9 percent.

Share of people at work who teleworked, August 2026

The all-occupations average hides a gap of nearly three to one between desk-and-code work and counter work.

Computer and mathematical		60.8%
Sales and related		23.9%
Office and administrative support		22.8%
All occupations		21.6%

So a Software Engineer should plan for the recruiter screen and probably the technical round to happen on camera, and treat an on-site as the exception rather than the default. A Bank Teller should plan for a branch, a manager, and a handshake — tellers, BLS notes, "usually work in branches of banks or credit unions."

There is a second split underneath the occupational one, and it points the same way. The American Time Use Survey found that in 2025, 51 percent of employed people with a bachelor's degree or higher did some or all of their work at home on days they worked, against 19 percent of those with a high school diploma and no college. The Occupational Outlook Handbook lists a bachelor's degree as the typical entry education for software developers and a high school diploma or equivalent for tellers. Two independent measures, one conclusion: the further a job sits from a physical counter, the more likely your first meeting with it happens through a lens.

One trap in reading these numbers: the industry rate and the occupation rate are different animals. Financial activities had the highest telework rate of any major industry in June 2026, at 52.5 percent, and finance and insurance was one of the two detailed industries BLS singled out as highest, at 59.0 percent. That number describes the analysts and the risk team. It does not describe the person at the window, who is counted in office and administrative support at 22.8 percent. If you are interviewing for a teller role at a bank whose head office is famously remote, the interview is still in the branch.

How dressed up is too dressed up?

Here is the position, and you are welcome to disagree with it: dressing two notches above the room is a worse mistake than dressing one notch below it. Underdressing slightly reads as ordinary. Overdressing dramatically reads as a misunderstanding of the job, and it is a misunderstanding the interviewer can see before you have said anything.

The reason is that clothes answer a question nobody asked out loud — have you met anyone who does this work? A candidate in a three-piece suit interviewing with three engineers in fleeces has answered no. The same suit in a bank branch, on the other hand, is fine, because that is roughly what the branch wears.

Watch what the job description implies rather than what the industry stereotype implies. O*NET's data on tellers is unusually blunt about the social shape of the role: 96 percent of respondents report constant contact with others, 91 percent have face-to-face discussions every day, and 65 percent call dealing with external customers extremely important. That is a job where a manager is hiring the person a nervous customer will hand a check to. Turning up looking like a visiting auditor is not a neutral choice.

What actually changes on camera

Less than people think, and not the things people focus on. The camera does not care about the cut of your jacket. It cares about contrast, pattern density, and where your head sits in the frame.

The same outfit, two formats

What to change when the interview moves to a webcam. Everything not listed here transfers unchanged.

	On camera	In the room
Color	Mid-tone solids; avoid bright white and pure black next to the face	Anything that suits you
Pattern	Nothing finer than a wide stripe — thin checks shimmer	Fine patterns are fine
Top half	Carries the whole impression; a collar or clean neckline reads well at low resolution	One part of a full silhouette
Bottom half	Still dress it — you will stand up	Obviously
Jewelry	Nothing that clicks against a desk or a lapel mic	Free choice
Layers	One you can remove; small rooms with a ring light get hot	Rooms are usually air conditioned
What the other side notices first	Framing, lighting, background	Handshake, posture, shoes

The last row is the one worth sitting with. In person, an interviewer's first impression is assembled from your whole body arriving in a room. On a call, they are looking at a rectangle roughly the size of a paperback, and the largest object in it is your face. That is why a mid-tone solid works: it puts a calm, even field behind the only thing being read.

Do I need a suit?

For a Software Engineer, no — a plain collared shirt or a good crew neck is the honest match to the room you would be joining. For a Bank Teller, yes, or close to it: a suit, or a blazer with a plain shirt, is what the branch looks like on a normal morning. When you cannot tell, a blazer over a simple top is the cheapest hedge in the whole business, because it dresses up when the room does and comes off when it doesn't.

Two jobs, two answers

The rule is the same for both example roles in this post. The output is completely different, and it is worth laying the two side by side to see why the format question comes before the clothing question.

The wages are the Occupational Outlook Handbook's May 2025 medians for those two detailed occupations, and the employment counts are its 2025 estimates; the telework rates are from the August 2026 household survey. The last two rows are not in any dataset — they are the inference you draw from the first four, and you should override them the moment a recruiter tells you something more specific.

One asymmetry the table does not show: the Handbook projects teller employment to fall 13 percent between 2025 and 2035, while the group software developers sit in grows 10 percent. That does not change what you wear. It changes who else is in the waiting area, and it is why the branch job rewards looking like you already work there more than the engineering job does.

A detail that changes the calculation for the engineer: among people in computer and mathematical occupations who teleworked in August 2026, an average of 76.9 percent of their hours were worked at home, against 67.5 percent across all occupations. Most of that group is hybrid rather than fully remote. Which means the two-video-rounds-then-a-visit pattern is common, and the visit is the round where the shoes matter.

What do you wear to the on-site after two video rounds?

More than you wore to the video rounds, and most candidates get this backwards. Three calls have gone well in a mid-tone shirt, so the shirt feels proven, and the fourth meeting is treated as the fourth episode of the same show. It is not. It is the first time anyone sees you below the sternum.

The on-site is the round with the widest audience. You will meet people who were not on any of the calls — a skip-level, someone from the team next door, whoever walks past the glass while you are at the whiteboard — and each of them forms an impression in a few seconds without the benefit of your answers. On a call, the frame does you a favor by cropping everything you did not think about. In a building, nothing is cropped.

So three things that were irrelevant on camera become relevant: shoes, hem length, and whether the whole outfit still works after forty minutes of sitting. Shoes are the one people skip, and they are the item an interviewer's eye reaches on the walk from reception. Clean and closed beats new and stiff, because you may be walking a floor.

For the Software Engineer, this is not an argument for a suit. It is an argument for the same shirt with trousers and shoes that were chosen rather than defaulted to, plus a layer, since a room booked for six people and used by three runs cold. For the Bank Teller the on-site is the first round, which is the whole point of working out the format before you work out the clothes — there is no proven shirt to fall back on.

Should you buy something for it?

Usually not, and the advice to "invest in a good interview suit" is worth pricing before you take it. BLS puts the median teller wage at $43,030 a year, or $20.69 an hour, as of May 2025, with the lowest 10 percent under $34,360. A suit bought on that wage for an interview that may not convert is a real cost, not a symbolic one, and the branch manager doing the hiring knows what the job pays.

What matters is that the clothes are clean, fit, and are not visibly borrowed from someone a size larger. A pressed plain shirt reads as competent at every wage level. If you are going to spend anything for a role likely to interview on camera — the Software Engineer case, where BLS projects 10 percent employment growth from 2025 to 2035 and about 106,100 openings a year across software developers, quality assurance analysts and testers together — spend it on a $20 lamp before a $300 jacket, because the lamp is in every round and the jacket is in one.

The camera is part of what you are wearing

An interviewer cannot separate your appearance from your setup, because both arrive in the same rectangle. A well-chosen shirt lit from below by a laptop screen still looks like a well-chosen shirt in a haunted house. Spend ten minutes on this once and it is fixed for every interview you do this year.

The ten-minute camera check

Do this once, the day before, in the clothes you plan to wear.

1. 1
   Put the light in front of you

   A window facing you, or a lamp behind the laptop. Never a window behind you — the camera exposes for the bright thing and turns you into an outline.

2. 2
   Raise the camera to eye level

   Books under the laptop. The default angle looks up your chin and puts the ceiling behind your head.

3. 3
   Frame from mid-chest up

   Head roughly a third of the way down the picture, a hand's width of space above your hair. Now your collar and shoulders are visible, which is the part of the outfit doing the work.

4. 4
   Test the shirt on camera, not in the mirror

   Record fifteen seconds and play it back. Fine patterns, logos and bright white all behave differently through a webcam than they do in a bedroom.

5. 5
   Check what is behind you

   A plain wall or a tidy shelf. If you use a virtual background, move your arms around and watch whether your shoulders dissolve.

Here is what the interviewer sees, in the two cases. Before, on a laptop on a low coffee table with a bright window behind you:

A dark shape against a white rectangle. The top of a head, cropped at the eyebrows. A voice, slightly delayed, coming from somewhere below the visible frame.

After, same shirt, same room, laptop on three books with the window in front of you:

A face, evenly lit, filling the middle third. Shoulders and collar visible. Eyes level with the lens, so you appear to be looking at the person rather than past them.

The candidate did not change. The second version is simply the one where an interviewer can spend the whole half hour listening instead of squinting.

Two more things belong in that ten minutes. Sit where a lapel mic or a headset cable is not going to drag across a necklace, and if you wear glasses, tilt the screen slightly so the reflected rectangle sits below your eyes rather than across them.

What this data does not tell you

None of these numbers are about interviews. BLS measures where employed people work, not where employers interview candidates, and no federal source records what anyone wore to one or whether it mattered. The American Time Use Survey found that 35 percent of employed people did some or all of their work at home on days they worked in 2025, and BLS reported in April 2026 that the monthly telework rate had moved only between 21.5 and 23.0 percent in the year through March 2026. August 2026's 21.6 percent sits inside that band. Stable, and still silent on your Thursday morning.

So treat the occupation figures as a prior, not a prediction. They tell you which assumption to make when the calendar invitation is ambiguous, and nothing more. One email to the recruiter beats all of it.

What to do this week

Ask the format explicitly when you accept the slot. "Is this a video call or on site, and how long?" gets you both the outfit and the travel time. Recruiters answer this in one line and think nothing of it.

Run the ten-minute camera check in the clothes you plan to wear, and save the setup — chair height, laptop stack, lamp position — so you are not rebuilding it at 8:50 on the morning of. If your first round is in a branch, the equivalent is trying the whole outfit on including shoes, a day early, in daylight.

Pick two outfits and stop deciding. One that works on camera, one that works in a room, both clean and both hanging up. Interview clothing is a solved problem the moment you stop re-solving it each time.

And if the harder problem is getting to the interview at all rather than what to wear to it, HeroApply fills the applications from your resume so the calendar invitations arrive faster than you can retype your work history.