Investor day logistics: the AV redundancy public companies need

The number I keep in my head is nine minutes. That’s how long a webcast dropped during a mid-cap company’s investor day two years ago, right as the CFO walked through revised full-year guidance. The sell-side analysts on the line saw a frozen slide and a spinning wheel while the CFO spoke to a number they couldn’t see. The stock did something ugly in the next hour. Nobody could prove causation, but nobody in that room ever again treated the webcast as an AV line item.

I plan convenings for a living, mostly associations and policy groups, and investor day is the one corporate event where I borrow the discipline of a live broadcast. The reason is simple. On this day the audiovisual system is a disclosure channel. If it fails while material information is on screen, you don’t have a production problem. You have a Regulation FD problem.

Map every single point of failure first

Before I talk about equipment, I draw the signal path on one page: laptop to switcher to screen, and laptop to encoder to CDN to the webcast player. Then I circle every box that, if it dies, takes down the whole thing. Each circle needs a twin. That’s the entire philosophy. Redundancy is not buying nicer gear. It’s making sure no single box can end the broadcast.

Here’s the map I walk, layer by layer.

Power: two feeds, one battery bridge

Everything hangs off power, so it goes first. I want the AV package on a dedicated circuit, and for the encoder and switcher I want a UPS battery bridge sized for at least 30 minutes, so a blip in building power doesn’t reboot the broadcast chain. At larger venues I ask about a house generator and whether the AV distro can tie to it. A conference center built for general sessions usually has the electrical capacity documented; an older ballroom at a hotel may not, and I find that out at the site visit, not on show day.

The slides: two computers, genlocked

The presentation runs from a primary laptop and a backup laptop, both loaded with the identical deck, advancing together, one live and one hot standby. If the primary freezes mid-guidance, the switcher cuts to the backup on the same slide with no black frame. This is cheap insurance, maybe $400 for the second machine and an operator’s attention, and it’s the failure the frozen-slide company didn’t cover.

The stream: two encoders, two paths out

This is the layer that fails publicly. I run redundant encoders and, where the venue allows, two separate internet paths: the venue’s hardwired fiber as primary and a bonded-cellular unit as backup, automatically failing over. A single hardwired drop is a single point of failure, and “the venue has great wifi” is not an answer for a webcast that regulators may later ask about. For the deeper reality of what actually breaks on hybrid connections, the hybrid event AV reality check is the piece I send clients who think one ethernet cable is enough.

I also insist on a hardwired internet connection for the broadcast that is physically separate from the guest wifi the in-room audience uses. When 200 people in the room open the webcast on their phones “to check the quality,” they can saturate a shared circuit and take down the very stream they’re checking.

Audio: the failure nobody notices until it’s fatal

Audio redundancy gets skipped because a dead mic seems minor next to a dead screen. It isn’t. The CFO’s lav should have a backup lav or a nearby handheld already open on the board, and the mix feeding the webcast should be a discrete send, not a microphone pointed at a speaker. I’ve seen a clean video stream carry unusable audio because the encoder was fed off the wrong bus. On investor day the words are the disclosure. Protect them like the slides.

The people layer: a producer whose only job is the broadcast

Gear redundancy fails without a human watching the right monitor. I staff a dedicated broadcast producer separate from the room’s AV lead, someone whose entire job is the encoder health, the CDN dashboard, and the standby paths. That person is not also running slides or wrangling the stage. When you hire this out, the how to hire a livestream production vendor guide covers the questions that separate a webcast vendor from an AV company that also owns a camera.

Rehearse the failure, not just the show

A dress rehearsal that only runs the happy path teaches you nothing. The day before, I have the team kill the primary encoder on purpose while we watch the player, to confirm failover actually happens and how many seconds it takes. We pull the primary laptop’s power to confirm the switcher cuts clean. We yank the fiber to confirm the cellular bond carries. If any of those tests reveals a gap, we still have a day to close it. Rehearsing the failure is the only rehearsal that matters here.

Venue selection follows the requirement

Because the AV requirement is this demanding, I let it drive the venue shortlist. A theater or performing arts space often comes with house power, rigging, and sightlines already solved, which shortens the load-in and lowers the redundancy build. A ballroom at hotels and resorts is flexible but you’re building the broadcast infrastructure from the truck up, so I price that reality in. The venue with the best chandelier is rarely the venue with the best pull-through for two internet paths.

The record and the archive

One more layer that’s easy to forget: capture a clean, redundant local recording of the whole session, separate from the stream, so the company has an authoritative archive of exactly what was said and shown. If a question ever arises about what was disclosed and when, that file is the answer. I store it off the show laptop, on two drives.

For the full sequence from save-the-date to the post-event archive, I build off the investor day event planning playbook and layer this redundancy map on top of it.

Investor day is the day your AV stops being production and becomes disclosure. Give every box a twin, staff a producer whose only job is the broadcast, and rehearse the failure before the CFO ever speaks.

Tell me your expected in-room headcount, your webcast audience size, and the date, and I’ll draw the single-point-of-failure map for your specific room.