To: [Engagement Manager] From: [Consultant] Subject: TTW composing costs — recommended next steps Composition drives 40% of hardback and 50-55% of paperback cost, and TTW is uncompetitive on simple work — but we still don't know whether that cost is actually excessive, because every title, simple or complex, runs through the same quality-control stages regardless of need. Two weeks of fact-finding point to a clear next step: run two parallel tests to find out where the money can come out — si...
$npx -y skills add welltraum/minto --skill raw --agent claude-code
Installs into .claude/skills of the current project.
Are you the author of Raw?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/welltraum-raw-0cc247fe)
To: [Engagement Manager]
From: [Consultant]
Subject: TTW composing costs — recommended next steps
Composition drives 40% of hardback and 50-55% of paperback cost, and TTW is uncompetitive on simple work — but we still don't know whether that cost is actually excessive, because every title, simple or complex, runs through the same quality-control stages regardless of need. Two weeks of fact-finding point to a clear next step: run two parallel tests to find out where the money can come out — simplifying quality control on simple titles, and improving compositor productivity.
1. Test whether quality control can be simplified on simple titles
- Select a sample of straightforward jobs and reduce or retime the checks they go through
- Measure the effect on quality and customer acceptance before rolling out
- Potential saving: up to 10% of total composing cost
2. Test whether a different composing method raises productivity
- Commission a methods study on manual composition, the tightest capacity constraint
- TTW currently trails the productivity benchmark by 20-50%; the study should identify where that gap can close
- George Kennedy has agreed to run this test
What I need now: sign-off to launch both tests, with Kennedy leading the productivity study.
---
Dropped from the source and not carried above, because none of it supports the recommendation directly:
- The list of people interviewed (Roy Walter, Brian Thompson, George Kennedy) beyond naming who will run the test.
- Staffing detail (below-market pay, two recent departures, late jobs, overtime 50%+ over budget, union claim) — real problems, but a separate issue from composing-cost reduction; worth its own note if you want it addressed.
- The suggestion to compare costs against three other printers — I don't think a simple comparison would settle the question, so I left it out rather than list it as a third workstream.
- The observation that managers disagree on whether costs are high — moot once the tests produce data.
