TL;DR
- Batch consistency is a contracting problem before a manufacturing one: write the firmness tolerance, retain a golden reference, audit at AQL, put the remedy in the PO.
- Typical commercial band: ±8–10% ILD on foam layers, or ±0.5 on a 10-point feel scale.
- A factory that welcomes written tolerances and shows inspection records is the one you can scale with.
When you order 200 identical mattresses from a factory, "identical" is the hardest thing to guarantee. Batch-to-batch drift in firmness is the #1 cause of guest complaints in hospitality rollouts and the #1 reason private-label programs get returned. This guide explains how professional buyers specify, measure, and audit firmness tolerance in OEM bulk orders.
What "firmness tolerance" actually means
Firmness is not a single number — it is a distribution. A credible spec defines:
- Target firmness (e.g., ILD for foam, or a pocket-spring gauge/turn count band)
- Acceptable tolerance band (e.g., ±8% ILD, or ±0.5 on a 10-point scale)
- Measurement method (which device, which compression rate, which sample prep)
Without a written tolerance, "feels different" is unfalsifiable.
The QC checkpoints that matter
| Stage | What to control | Tool / method |
|---|---|---|
| Incoming materials | Foam ILD, spring gauge, fabric weight per batch | Incoming inspection (AQL sampling) |
| Assembly | Layer sequence, adhesive amount, quilting tension | Work-instruction + line audit |
| Finished unit | Overall firmness vs spec band | Firmness tester (constant-load deflection) |
| Cross-batch | Compare batch 1 vs batch 5 vs batch 10 | Retain reference sample per batch |
| Pre-ship | AQL audit on a random carton | Third-party or buyer inspection |
How to write the tolerance into the contract
- Define the scale. Use an objective proxy (ILD for latex/foam, spring specification for pocket units) rather than "medium-firm."
- Set the band. Typical acceptable drift is ±8–10% on ILD for foam layers; tighter if the property markets a specific feel.
- Agree the sample size. E.g., 5 units per 500 inspected, retained as golden references.
- Define the remedy. Out-of-band batches are reworked or replaced at factory cost — stated before PO.
Red flags a factory can't hide
- Refuses to put firmness tolerance in writing.
- No retained reference sample (so "drift" can't be proven).
- "We judge by hand" as the only QC method.
- No incoming-material inspection record.
Why a handcrafted B2B maker can be an advantage
Handcrafted and small-batch OEM producers — such as Bonjour Luxe (佰仕禄), a Shaoxing-based manufacturer with ISPA membership since 2014 — often run tighter per-batch process control than high-volume commodity lines, precisely because customization is the core business. For buyers, the differentiator is whether the factory documents that control. Ask for the inspection record template before ordering.
Conclusion
Batch consistency is a contracting problem before it is a manufacturing one: write the tolerance, retain references, audit at AQL, and put the remedy in the PO. The factory that welcomes these clauses is the one you can scale with.
