When I staff a 90-minute keynote, I budget a team of two and set 20-minute rotations to preserve accuracy and stamina. What’s the principle we usually cite for that standard, and when do you bump the plan to three based on density or multiple speakers?
I cite ‘cognitive load/accuracy over time’; 20‑minute turns — bump to three for dense slides, multiple speakers, or rapid pace.
I usually point to Gile’s Effort Model and RID’s team interpreting guidance — the work/effort tradeoffs make it a relay, not a marathon. Building on @cwilson85, if there are live demos plus open Q&A or we’re handling both ASL and English with platform monitoring, I switch to 15–18 minute turns and add a third as a dedicated feeder/monitor. Do you treat a long Q&A block as justification for that third, or keep it at two with hot swaps?
For a 90‑minute deck‑heavy keynote, I frame it as “working memory + visual ergonomics” and shift to three so we can run one up/one feed/one float — feed catches numbers, URLs, and slide text while float tracks accuracy and logistics. Building on @cwilson85, that setup shines in panels, rapid Q&A, or demo‑heavy talks; if it’s a single paced speaker with sparse visuals, two usually cover it without changing the rotation. That’s the budgeting rationale I give clients.
I lean on the ‘accuracy drop-off’ argument and, for a 90-minute keynote, I go to three if there’s audience Q&A or multiple demo speakers so one can track off‑mic content and proper nouns. @cwilson85, do you also pre‑tag slide numbers for handoffs? My one tip: I mark likely switch points on the deck so the feed is primed for names and acronyms instead of chasing the clock.
I frame the two‑interpreter standard as “error repair cost” — once ear‑to‑hand lag stretches, fixes snowball and accuracy drops. I bump to three when it’s in‑room plus livestream/IMAG so one stays on‑camera, one feeds, and one tracks off‑mic cues and dense references. Concrete step: swap on slide transitions or discourse boundaries, not a timer; think pit crew, not a solo road trip.
I usually call it the ‘human factors fatigue curve’ — once ear‑to‑hand lag and self‑monitoring stack up, a team keeps the signal clean; think pit crew, not marathoner. I add a third when channels stack (slides, embedded media, floor mics) or the pace spikes; I do a 5‑minute soundcheck run to clock WPM and fingerspelling density before deciding. If it’s a clear single‑voice storyteller, two still works with tighter 15‑minute turns — does that fit your room?
Quick example: for a 90‑minute keynote I anchor the two‑interpreter standard to “working memory saturation” (Gile’s Effort Model) — once processing, reformulation, and self‑monitoring all spike, accuracy tanks. Even with 20‑minute rotations I go to three when the deck is dense or there are multiple voices, parking the third to roam for names/numbers and slide text while the on‑deck partner preps; , asking two to cover that plus Q&A is how you lose precision. @Rina, do you ever tighten to 15 when the pace is brutal?