Do I need to be on camera?

No.

You do not need to be on camera. You need to be findable, credible, and consistent, and the camera is one route to that rather than the toll gate it gets described as.

The question underneath it is simpler. Which format will you still be running in a year?

That question has a measurable answer, and it is not encouraging.

Most of what gets launched is abandoned within months

Podchaser pulled every podcast that debuted in the first half of 2026, all 153,767 of them, and checked which ones were still publishing on June 30. Roughly 42% had already gone quiet, inside six months of starting.

That number is about podcasts, and the pattern is about people. A format you dread produces three strong pieces and then a gap, and the gap turns into an archive with a date on it that tells every visitor exactly when you gave up. An abandoned channel costs more than an absent one, because the absence says nothing while the abandonment says something specific.

So the format question is a stamina question. What you can do for 50 weeks beats what performs best in a benchmark table.

Video does earn trust faster, and the evidence is narrower than the claim

You will read that video builds trust. It has research behind it, and the research is worth knowing precisely, because the precision is what tells you when it applies.

The cleanest study is old and small. Researchers at Michigan put 45 three-person groups through a cooperation game, letting each group communicate face-to-face, over video, over audio, or by text chat. Video groups reached the same level of trust as the groups sitting in the same room. Audio landed in between. Text chat did not get there at all. Trust in the mediated conditions also formed more slowly, and the researchers described it as fragile.

That is a laboratory game among strangers, not a study of how anyone buys legal advice. It supports one thing: seeing a face while someone speaks moves trust faster than reading their words. It does not support the idea that a face on camera sells professional services, and nobody has shown that.

Now notice who publishes the rest of the numbers you have seen. The statistics claiming that most buyers prefer video, trust video, and shortlist from video come from video software companies, video agencies, and the tools that sell you an easier way to film. Even the platform data disagrees with itself. One 2026 benchmark set has video ahead of text posts on engagement, while another reads the same platform and reports video views falling sharply year over year, with written posts pulling ahead. When measurement disagrees this loudly, the conclusion worth drawing is that format is not the variable carrying most of the outcome.

Your audience is doing both

Podcast listeners settled this question for themselves. Edison Research's 2026 study of the podcast audience found that most people who consume podcasts both listen and watch, about one in five consume audio only, and a small sliver watch without ever listening.

Read that as permission. Audio-only work reaches a durable audience. Written work reaches a different one, at a different depth, and it is the version an AI system can quote back to a buyer who asked about your field. Video reaches the discovery layer, which is where people who have never heard of you are.

None of those is optional in the sense that a missing one ends you. Each is a door. You are choosing which doors to open and how many you can hold open at once.

What the camera does for a buyer

Someone hears your name, and within a day they look you up. That moment decides more than any single channel you pick.

In that lookup, a face does specific work. It confirms that a person stands behind the writing. It shows how you handle a question, which is the thing your buyer cannot get from a case study. And it settles the small doubt that follows every polished page in 2026, which is whether a person wrote it.

That last one is new. As text gets cheaper to generate, the value of visible authorship rises. Your buyer does not think this consciously. They just find the page more convincing when they can see who is talking.

There is a cost on the other side, and it depends on your field. Some professionals watch peers post explainer videos and read the whole genre as cheapening. If your buyers hold that view, a weekly talking-head clip works against you, and a recorded conversation with a peer does not. The format carries a register, and the register has to match the work.

An avatar reading your script is not you on camera

The 2026 shortcut deserves a straight answer. You can now record two minutes of yourself, train an avatar, and publish videos of your own likeness saying words you typed and never spoke. The good ones pass a casual look.

I would not do it, and I would tell you not to.

You sell judgment. The delivery is part of the proof, because the way you handle a question is the sample your buyer is evaluating. An avatar removes the sample and keeps the face, which is the opposite of what a considered buyer wants. Several jurisdictions now require disclosure when a synthetic presenter appears in advertising, and a disclosure line under your face is not a neutral thing for a practice selling trust.

Voice cloning for translation into a language you do not speak is a different question with better answers. Reading a script you never spoke, in a rendered copy of your face, is not.

Choose the format you would still run on a bad week

Three shapes hold up for expert practices, and you can run one well or two adequately.

Writing compounds hardest and asks the least of your nerve. It is searchable, quotable, and the version that ends up cited when someone asks a chatbot about your field. Audio conversation asks little preparation and produces the most raw material per hour, which is why interviews suit people who freeze in front of a lens and speak freely with a peer. Camera work at low frequency, one recorded session a month rather than daily clips, gives you the trust the research supports without the treadmill that ends most of the shows launched this year.

The pattern worth noticing in all three: answering is easier than presenting. Almost nobody who dislikes the camera dislikes being asked a good question. If the camera is the problem, get out of the presenter's chair and into the interview, either as the guest or by running the conversation yourself.

Whatever you choose, one voice has to run through it. The writing, the audio, and the video should sound like the same person, because they are, and because a buyer who meets two of them will notice if one of them is a stranger.

Tonight, record five minutes on your phone answering the question a client asked you this week. Watch it once. Then decide whether you would be willing to do that 50 more times, and pick your format from the answer rather than from what the internet says the algorithm wants.