Video SOPs: A 5-Step Method to Capture a Retiring Expert's Know-How
D-SKT-04 | First published 2026-07-18 | Revised 2026-07-18 | Primary sources: 3
Disclosure: Gemba Works Lab is supported by advertising. This page contains no affiliate links and recommends no specific products; if affiliate links are added in the future they will be clearly marked (per the FTC Endorsement Guides).
Written SOPs capture what to do. Video captures the part writing loses: hand position, rhythm, the sound of a cut going right, where the expert's eyes go. For a small plant facing a retirement, video is also the cheapest serious capture tool available — the camera in your pocket is sufficient, and the method below needs no editing skills beyond trimming clips. This guide gives a supervisor or trainer a 5-step method that turns roughly one day of a veteran's time per job into a training asset a new hire can actually learn from. It targets the jobs your skills matrix flagged as one-deep.
KEY POINTS
The core is a two-pass recording: film the hands doing the job, then film the veteran watching the footage and explaining what they were sensing. Pass 2 is where the tacit knowledge surfaces.
Chapters beat productions. Ten 2-minute clips indexed by step outperform one polished 40-minute film — trainees rewatch steps, not movies.
A video SOP is only "done" when a trainee has used it to perform the job to standard. Verification, not filming, is the finish line.
Why this is worth a day per job
The labor math justifies the effort. BLS projects about 34,200 annual openings for machinists and tool and die makers through 2034 — overwhelmingly replacement demand as experienced workers retire or leave, against slightly declining total employment (source: BLS OOH — Machinists and Tool and Die Makers). Industry-wide, a Big Four consultancy and The Manufacturing Institute project up to 1.9 million manufacturing jobs could sit unfilled by 2033 (source: The Manufacturing Institute & a Big Four consultancy, 2024). If replacements are scarce, the leverage move is making the replacements you can get productive faster — which is precisely what indexed video of your best practitioner does. One day of veteran time per critical job is cheap against a multi-month ramp-up gap; our costing guide shows how to put your own numbers on that comparison.
The 5 steps
Figure: The 5-step video SOP method. Pass 2 (the why-interview) is where tacit knowledge becomes explicit.Source: Gemba Works editorial team; two-pass structure per the externalization stage of the SECI model (Nonaka & Takeuchi).
Pick and scope the job. Take a one-deep job from the skills matrix. Scope it to one setup or one operation — "everything Joe knows" is not a scope; "first-article setup on the Mazak for thin-wall aluminum" is.
Pass 1 — film the hands. Real production job, no rehearsal. One person films (phone, landscape, close on hands and workpiece); the veteran works normally and narrates only if it comes naturally. If the job includes safety-critical steps — lockout, guarding checks — film them explicitly and in full; a video SOP that skips them teaches the wrong lesson (OSHA requires training to be understandable to workers; see osha.gov for your obligations).
Pass 2 — the why-interview. Same day or next: sit the veteran in front of the footage with the question list below, and record that conversation. Every pause, every unexplained adjustment is a question site. This pass regularly doubles the useful content — it is the step almost everyone skips, and the reason most shop videos teach motions but not judgment (the "externalization" stage described in our tacit-knowledge guide).
Cut into chapters. Trim into clips of roughly 1–3 minutes, one step each, named plainly ("03 — indicating the part," "07 — what chatter sounds like"). Add a one-page index with links/QR codes and the tolerances that pass 2 surfaced. Store where production can reach it, with a named owner for updates.
Verify with a trainee. A learner performs the job using only the chapters plus normal supervision. Where they stumble, the video is missing something: film the gap, re-cut, retest. Done means demonstrated.
The why-interview question list
Ask at every decision point in the footage
"What were you looking at right there?"
"What would you have done if it looked/sounded/felt wrong?"
"What do beginners always get wrong at this step?"
"You adjusted something just now — what triggered that?"
"Which suppliers/materials/machines change how you do this step?"
"What's the one thing you'd check if the part came back bad?"
"When is it NOT okay to proceed, even if everything looks fine?"
Use a curious junior as interviewer, not a manager — the veteran explains more to someone who genuinely doesn't know, and OSHA's worker-participation guidance points at the same dynamic: front-line knowledge surfaces in settings where sharing it is safe and valued (source: OSHA — Worker Participation).
Pitfalls that waste the footage
Filming a demonstration instead of a job. Staged runs omit the recoveries and workarounds — which are the valuable part.
Skipping pass 2. Hands-only footage looks complete and teaches half the skill.
One long video. Nobody scrubs through 40 minutes to rewatch one step; unchaptered footage goes unwatched.
No owner, no updates. When the method changes and the video doesn't, trust in the whole library dies. Assign an owner, and treat method changes the way a kaizen standard update is treated.
Framing it as extraction. "Record everything before Joe leaves" insults Joe. Frame it as building the training library that carries his standards forward — veterans respond to legacy, not to being downloaded.
Last updated: 2026-07-18 | Primary sources referenced: 3 | This guide is general training advice, not safety compliance advice; your OSHA training obligations are defined at osha.gov. Spotted an error? Tell us via the contact page.