Prompt Details
Model
(claude-5-opus)
Token size
628
Example input
[SOURCE]: From "Focus Time in Hybrid Teams" (Kellerman & Osei, Journal of Organizational Behaviour, 2025). The authors tracked calendar and IDE telemetry for 1,240 knowledge workers across 14 firms over nine months. Teams that adopted a "no-meeting Wednesday" saw uninterrupted work blocks rise from a median of 47 minutes to 96 minutes. Self-reported burnout fell 8% but the confidence interval crosses zero, so the authors describe the effect as "suggestive rather than established". Output measures were mixed: story points closed rose 11% in engineering teams and were flat in design teams. The authors caution that firms which volunteered for the study were already meeting-averse, and that the telemetry cannot distinguish deep work from idle IDE sessions. They recommend policy be set at team level rather than company level, noting three firms where a company-wide ban simply pushed meetings into Tuesday and Thursday, producing what one manager called "a wall".
[WHY]: We're deciding whether to make no-meeting Wednesday company-wide next quarter, and I'll have to argue it either way in the leadership review.
[SYSTEM]: (none provided)
Example output
# No-meeting Wednesday roughly doubles focus blocks, but a company-wide ban can just relocate the meetings
## When you'll want this
- You're in the leadership review next quarter and someone says "show me the evidence that this works" — or "show me why we shouldn't just mandate it everywhere."
- Someone cites this paper as proof that focus policies raise productivity, and you need to say precisely what it does and doesn't support.
- A team wants to opt out of the policy, or keep their own day instead of Wednesday, and you have to decide whether uniformity actually buys anything.
- Six months in, you're deciding what to measure to tell whether the policy is working.
## The points
**1. In three of the studied firms, a company-wide ban pushed meetings into Tuesday and Thursday rather than removing them.** One manager described the result as "a wall." Kellerman & Osei recommend policy be set "at team level rather than company level." *(my read: this is the single most load-bearing finding for our decision, and it cuts against the company-wide version we're proposing. If we go company-wide anyway, the honest framing is that we're doing it for coordination reasons, not because the evidence supports it.)*
**2. Median uninterrupted work blocks rose from 47 minutes to 96 minutes in teams that adopted no-meeting Wednesday.** This is the study's cleanest and largest effect. *(my read: this is the number to lead with — it's a behavioural change, directly measured, not self-reported.)*
**3. Output effects were mixed by function: story points closed rose 11% in engineering teams and were flat in design teams.** The authors report this as "mixed," not as a win. *(my read: expect this to be the first thing a skeptical exec pulls out. Better to raise it ourselves than be caught with it.)*
**4. Self-reported burnout fell 8%, but the confidence interval crosses zero — the authors call the effect "suggestive rather than established."** Do not upgrade this in the review. If we present it as a wellbeing win we are overstating what the paper says, and it's the easiest claim for anyone who reads the abstract to knock down.
**5. Participating firms volunteered and were "already meeting-averse," which the authors flag as a limit on generalising.** *(my read: our meeting load is probably heavier than the study sample's, which cuts both ways — more headroom for gains, but also more pressure for meetings to reappear somewhere else, i.e. more "wall" risk.)*
**6. The telemetry "cannot distinguish deep work from idle IDE sessions."** So the 47→96 minute figure measures uninterrupted time-at-tool, not uninterrupted thinking. *(my read: this is why our own success metric shouldn't be a focus-time dashboard — it would be trivially gameable and wouldn't mean what it looks like.)*
**7. Study spec, for citation:** calendar and IDE telemetry, 1,240 knowledge workers, 14 firms, nine months. Kellerman & Osei, *Journal of Organizational Behaviour*, 2025 (volume, issue, page numbers and authors' first names — check source).
## What this note leaves out
Nothing material from the excerpt I read — but the excerpt is itself a summary, so it omits the paper's actual statistics: effect sizes, confidence intervals, controls, how "uninterrupted block" was operationalised, the composition of the 14 firms, and any explanation the authors offer for why design teams saw no output change. Go to the paper before quoting numbers in a slide.
## Filing
New vocabulary, since none was supplied — small set, adjust to fit:
Tags: `#evidence-note` `#meetings` `#focus-time` `#hybrid-work` `#decision-input`
Links: `[[No-meeting Wednesday decision — Q4 leadership review]]` · `[[Meeting load policy]]` · `[[Team-level vs company-level policy]]` · `[[Metrics we should not use]]`
## The thread worth pulling
The paper reports where meetings moved, but not whether total meeting hours fell anywhere. Did no-meeting Wednesday reduce meeting volume, or only redistribute it — and in the teams where focus blocks doubled, what got cancelled outright? That's the number that would actually settle our decision, and this source doesn't have it.
Get this prompt + 9 more with Select$14 your first month, then $19/mo$14 first month, then $19/mo
View SelectBy purchasing this prompt, you agree to our terms of service
CLAUDE-5-OPUS
Most notes fail at retrieval, not at capture. Paste an article, transcript, meeting or highlights and it writes the note backwards from the day you'll need it: a title that states the claim, the future moments this note is for, atomic points that stand alone, and a hard line between what the source said (attributed, in its words) and what you inferred, marked (my read). It says what it dropped, fits your tags and links, and never hardens a hedge or guesses a figure. Tested on Claude Opus 5.
...more
Added over 1 month ago
