Digital product2026

SYNAPSE

AI research tool

synapse.app/r/four-day-week

Research / Four-day week — manufacturing8 sources · 64 claims · 5 open questions⌘K

Sources 8

S01Pilot report — Nordhavn Components

S02Shift compression and defect rates

S03Employer survey, Baltic manufacturing council

S04Overtime substitution in reduced-hours plants

S05Interview — plant director, Halden Foundry

S06Energy cost per unit under 4×10 scheduling

S07Union position paper on schedule reform

S08Maintenance windows and line availability

Topics

ProductivityOvertimeQualityEnergyRetentionScheduling

Question 03

Does a four-day week reduce output in mid-size manufacturing?

01 Read 8 sources · extracted 64 claims · grouped into 6 topics

The evidence separates two outcomes that are often merged. Output per worker-hour rises in every pilot that reports it S01S03S07, while total weekly output depends on how the fifth day is covered S04S08.

Where plants replaced the removed day with overtime, the gain disappears within two quartersS04S05. Plants that closed the line for a full day kept it, and also report lower energy cost per unit S06.

Gap No source measures output beyond 18 months. 2 of 8 sources are employer-reported.

Ask about these sources, or challenge a claim…↵

Claims in this answer

C0140.78

Output per worker-hour rose in the first two quarters after the change.

▲ 3▼ 1S01 S03 S07 S04

C0150.61

Total weekly output stayed within ±3% of baseline.

▲ 2▼ 2S01 S03 S04 S08

C0190.72

Gains fade once overtime is used to cover the fifth day.

▲ 2▼ 0S04 S05

C0220.44

Defect rates fall when shifts are shortened, not compressed.

▲ 1▼ 1S02 S06

Area: Digital product · 2026
Role: Product strategy, UX architecture, design, interface development
Scope: 11 screens: workspace, knowledge graph, reader, comparison, report, mobile

Task

AI chats answer single questions well but hide the path from source to sentence — and that path is what an analyst has to defend.

Key decision

The unit of the product is not the message. It is the claim — a statement with an identity, a confidence and a page number.

Proof 01

Knowledge graph

Sources, claims and topics form one graph. Solid edges support, dashed edges contradict. A selected contested claim explains why it is contested.

What it showsData visualisation and complex interaction.

synapse.app/r/four-day-week/graph

Research / Four-day week — manufacturing8 sources · 64 claims · 5 open questions⌘K

ProductivityOvertimeEnergyC014C015C019C027C022S01S03S04S05S06S07S08S02NordhavnHalden

SourceClaimTopicEntity supports contradicts

AllContested onlyBy topicTimeline

Proof 02

Document reader

Source text set in serif on paper, extracted claims in the margin. A contradiction is flagged where it occurs.

What it showsTypography and readability inside a working tool.

synapse.app/s/S04

Research / Four-day week — manufacturing8 sources · 64 claims · 5 open questions⌘K

Page 9 of 31

3. Substitution effect

Across the eleven plants in the sample, the removal of the fifth production day was followed by a rise in scheduled overtime within the first quarter. Where overtime was capped by agreement, weekly output fell by 6.4% against the prior-year baseline, while uncapped plants recovered the volume at a higher labour cost per unit.

We note that the baseline year coincided with unusually strong order books. Measured against a three-year average the decline narrows to 2.1%, which falls inside the range reported by employer surveys.

The effect is therefore better described as a substitution of paid overtime for regular hours than as a loss of productive capacity. Output per worker-hour improved in nine of eleven plants.

Proof 03

Report builder

Coverage is measured: a sentence without a source is flagged before a reader finds it.

What it showsProduct thinking: an interface that prevents the mistake.

synapse.app/r/four-day-week/report

Research / Four-day week — manufacturing8 sources · 64 claims · 5 open questions⌘K

Draft 4 · edited 12 min ago

1 Productivity and output

Hourly productivity improves consistently after the change S01S03S07. Total output is stable where the removed day is not replaced with overtime S01, and declines modestly where it is S04.

Uncited “Most plants adapt within one quarter.” — no source supports this sentence. Find evidence · Remove

The apparent contradiction between pilot reports and S04 is explained by baseline choice; both positions are shown in Table 2 C015.

Table 2BaselineΔ weekly output

S01 Nordhavn pilotPrior year−1.2%

S03 Council surveySelf-reported±0

S04 Eleven plantsPrior year / 3-yr avg−6.4% / −2.1%

What was made

Product object model
UX architecture for five views
11 interface screens
Knowledge graph interaction model
Mobile companion
Visual system and component set

SYNAPSE

Sources 8

Topics

Does a four-day week reduce output in mid-size manufacturing?

Claims in this answer

Knowledge graph

Document reader

3. Substitution effect

Report builder

1 Productivity and output

Four-day week — manufacturing

Since yesterday

Total weekly output stayed within ±3% of baseline.

Supports

Contradicts

Add to: Four-day week