On our last 12-week interiors project, we protected two 90-minute Revit QA sessions each week (Tue/Thu at 8 a.m.) and still hit permit submittal by day 60 by front-loading coordination in the pull plan. How are you building recurring training into the schedule without burning float or denting client satisfaction?
And we treat training as a ‘clinic’ right after Monday standup: 20 minutes, one live fix from last week’s issues log, recorded to a 5‑min Loom that’s assigned before related tasks — like brushing teeth, not a root canal. Because it’s tied to the board and WIP limits, float stays intact; the only caveat is enforcing the timer or it balloons. Would you trade one coordination huddle per sprint for a QA‑lite if it’s fed by the pull‑plan constraints?
We gate training with a simple rule: “lab runs only if last week’s PPC ≥ 85%; otherwise it flips to production support,” which keeps float intact while still holding a 25‑min Thursday live lab tied to next week’s scope plus a 3‑min Loom as the pre‑task check. Building on @james_r89’s clinic idea, we rotate the host so whoever fixed the issue teaches, which seems to stick better. Have you tried a quick “model floss” minute at the end to clear tiny tech debt before it calcifies?
We turned training into a ‘teach-back’ tied to each sheet/family’s first internal review: the author gets 10 minutes to show how they built it, and the output is a checklist tweak or reusable detail, so it rides the same calendar block and doesn’t touch float. If a milestone risk goes yellow, the teach-back auto-defers and the item just gets reviewed — borrowing @leonardw317’s gate idea. It’s like seasoning while you cook, not scheduling a cooking class; have you tried logging these as DoD deliverables?
On our last interiors run, we turned the Tue 8 a.m. Revit QA into a ‘live driver’ rotation: one person takes 7 minutes to screen-share the audit view and walk through fixing two flags, then we switch straight back to QA so the 90-minute block stays intact and we still hit “day 60”. Small caveat: only record and file the clip when the pattern repeats; otherwise the admin overhead isn’t worth it.
We timebox a 15-minute pre-publish ‘pit stop’ right after standup: the day’s publisher pairs with a buddy, runs an 8-check preflight, and records a 90-second Loom that becomes the micro-training. It only works if you enforce “no meeting without an artifact” and keep it to 15 or it balloons. Anyone tracking preflight misses vs RFIs to prove it’s earning its slot?