What a corporate livestream actually costs, and what it returns

Most people asking what a corporate livestream costs are really asking a different question: whether the cheap version is good enough. Usually it is not, and the reason is narrower than you would think. It is almost never the cameras. It is the audio, the redundancy, and whether anyone is watching the stream while it is running.

Here is what actually moves the number.

The five things that set the price

Camera count. One camera is a recording. Two is a conversation. Three or more is a broadcast, because now a director is cutting between angles and someone has to be doing that in real time. Each additional camera adds an operator or a PTZ, a cable run, and a line on the switcher. This is the most visible cost driver and the one clients focus on, which is why it is first and also why it is not the most important.

Audio. This is the one that decides whether people stay. Viewers will tolerate a soft image for an hour. They will leave in ninety seconds over bad sound, and they will not tell you why. A room mic picking up HVAC and chair scrapes is the single most common reason a corporate stream underperforms. Proper audio means lavaliers or a board feed off the house system, a mixer, and someone riding levels.

Redundancy. A second encoder, a bonded cellular uplink alongside the hardwired connection, a backup recording running locally. None of it is visible when the event goes well. All of it is the entire budget when the venue's internet drops during the CEO's opening remarks.

Remote presenters. Bringing someone in over Zoom or Teams and having them look like they belong in the broadcast is a pipeline, not a checkbox. Audio sync, framing, a clean background, and lower thirds that match the in-room graphics. Two remote speakers is a different production than zero.

Graphics and post. Lower thirds, a title package, a countdown, sponsor bumpers. Then afterward: the edited recording, the highlight cuts, the individual session files. This is where scope quietly doubles if nobody wrote it down in advance.

The bandwidth question, with actual numbers

A 1080p stream at a decent bitrate wants about 6 to 8 Mbps of upload, sustained. Add a second bitrate ladder for adaptive delivery and you are near 12. Add a backup encoder pushing simultaneously and you are at 20.

That is why we ask for 20 Mbps dedicated upload, and why "we have great WiFi" is not an answer to the question.

Two things to know about venue internet. First, the number the venue gives you is the number on their bill, not the number available in your ballroom at 2pm with 400 phones attached to the same access points. Second, "dedicated" has to mean a hard line to your encoder, not a VLAN on the guest network. We test it during the site survey because we have been wrong about this before by trusting the answer.

Working out whether it pays

Skip the engagement metrics. Ask three questions.

What did the room cost? If you are flying forty people in for a two-day meeting, the stream is not competing with zero. It is competing with the airfare, the hotel nights, and two days of everyone's time. A hybrid format that lets half of them attend from their desk pays for the production before you count anything else.

What is the recording worth? Most corporate events get produced once and watched once. That is a waste. The same production, recorded properly, becomes onboarding material, sales enablement, the internal announcement people missed, and a highlight reel. The marginal cost of a good recording at a shoot you are already doing is small. The marginal value is not.

Who could not be in the room? For a town hall, this is the whole point. The people who missed it are the ones the message was for.

What I would actually cut

If the budget is tight, cut cameras before you cut audio, and cut graphics before you cut redundancy.

A clean two-camera stream with good sound and a backup uplink beats a four-camera stream with a room mic and a single point of failure. Every time, and it is not close. The second one looks better in a proposal and worse on the day.

The other thing I would not cut is somebody watching the stream while it runs. It sounds absurd to pay a person to look at a monitor. But an encoder can be publishing a black frame to your audience while the room looks perfect, and the only way anyone finds out is if a viewer emails someone, which happens about twenty minutes in. We have caught this on other people's productions.

What to send us

If you want a real number rather than a range, we need three things: the date, the venue, and how many people are speaking remotely. That is enough to scope it. Everything else we can figure out on the call.