You have 4 hours of deposition footage. You get 4 minutes at trial.
That's the math every trial team faces. A deposition runs for hours, but jurors will see a handful of clips, maybe less.
Pick the wrong ones and you waste your best shot at the witness. Pick the right ones and you land the point you actually need.
Most attorneys choose clips by memory and gut feel. That's a problem, because gut feel doesn't account for how a clip actually plays to someone who wasn't in the room.
Why instinct fails on deposition clips
You were there. You know the context, the tone, the pause before the answer.
The jury wasn't. They see 20 seconds of a stranger on a screen, and they judge that stranger fast.
Research on facial expression and credibility judgments shows viewers form snap impressions of honesty and trustworthiness within the first few seconds of watching someone speak (Willis & Todorov, 2006, Psychological Science). Deposition clips are no different. A witness who looks confident to you on paper might look shifty on screen, and you won't know until it's too late.
The clips that read wrong on video
Some patterns show up again and again once you test deposition footage on real viewers.
A witness who pauses to think reads as honest in person but hesitant on video. A lawyer's leading question can make an answer sound coached, even if the witness believed every word. Long, technical answers lose attention fast, and once a viewer checks out, the rest of the clip is wasted air.
None of these problems show up when you're reading a transcript. They only show up when someone watches the tape.
How facial coding fits into clip selection
EmotionTrac runs deposition clips past opt-in panelists and reads their facial expressions frame by frame using facial action coding (FACS). The result is a second-by-second emotional read on every candidate clip, before you commit to it in front of a jury.
You're not guessing which 30 seconds land. You're watching where attention holds, where confusion spikes, and where trust in the witness rises or drops.
What the data actually shows you
Four signals matter most for deposition clips:
Attention: does the viewer stay locked in, or does focus drift after the first 10 seconds?
Credibility: does the witness's expression match the tone of the answer, or is there a mismatch that reads as evasive?
Confusion: does a legal or technical term cause a visible drop in comprehension?
Emotional shift: does the panelist's reaction change mid-clip, and if so, at what exact second?
A practical workflow for trial teams
Step 1: shortlist your candidates
Pull every clip you're considering, even the ones you're only half sure about. Aim for 8 to 12 short segments rather than 2 or 3 long ones. Short clips are easier to test and easier to compare.
Step 2: run the panel
Send the clips to an opt-in panel that matches your likely juror demographic as closely as possible. Age, region, and general education level all affect how a clip reads, so match the panel to the venue.
Step 3: read the second-by-second data
Look for the moment attention or trust drops. If it happens in the first 5 seconds, the clip probably opens on the wrong beat. Trim it and retest.
Step 4: build the final montage
Once you know which clips hold attention and which support credibility, sequence them. Order matters. A strong clip followed by a weak one can drag the whole sequence down in a viewer's mind, even if the weak clip is short.
Where this fits your trial prep timeline
Run this process 3 to 4 weeks before trial, not the week before. That gives you time to re-cut clips, retest a revised sequence, and brief co-counsel on what the data actually shows.
It also gives you leverage in settlement talks. If your deposition clips test as strong on credibility and attention, that's a data point worth mentioning across the table.
If they test weak, you've just saved yourself from finding that out live, in front of a jury, with no way to take it back.
The bottom line
Deposition footage is evidence, but how you cut it is a decision. Facial coding turns that decision from a guess into a read you can defend.
You don't need every clip to be perfect. You need to know, before trial day, which ones actually work on the people watching them.
Try it: Schedule an EmotionTrac demo and see second-by-second emotion tracking in action. Or visit Insights for more information.
Sources
Willis, J., & Todorov, A. (2006). First impressions: Making up your mind after a 100-ms exposure to a face. Psychological Science, 17(7), 592-598.