Methodology: How We Decide What to Decode
Brands rarely publish batch code formats, so most “decoders” online are educated guesses. We try to separate what is documented from what is not.
Evidence levels
| Level | Meaning | How we treat it |
|---|---|---|
| A | Published by the manufacturer | Would be decoded with confidence. None reached yet. |
| B | Several independent, dated confirmations | Would be decoded with confidence. None reached yet. |
| C – Limited | Documented by third-party sources, with some mappings or dated examples | Decoded, labelled as limited evidence |
| D – Weak | Single, vague, conflicting or inconsistent claims | Not decoded |
Our rules
- Document first. We decode only formats that are written down with a mapping.
- No extrapolation. If a source maps some year letters, we do not invent the rest.
- Show ambiguity. One-digit years repeat every ten years, so we show both possibilities.
- Conflicts mean no result. When sources contradict each other, we decode nothing.
- Codes give manufacture dates only. A “typical use-by” adds a general 30-month guideline and is labelled as such. The printed date and PAO always win.
- Synthetic examples are not evidence. Example codes that look like templates are not used.
The research so far
| Unique brands listed | 334 |
|---|---|
| Brands with a documented rule | 4 |
| Brands with partial information | 24 |
| Brands still unresolved | 306 |
| Sources reviewed | 22 |
| Distinct source-dated examples | 7 |
| Rules independently tested | 0 |
| Product-specific records | 0 |
Advertisement
The research is not exhaustive. No source we found is an official manufacturer document, which is why no rule reaches level A or B. Last researched: 2026-10-12. See the source list.
Help us move rules up
The fastest way to add or upgrade a rule is a real product photo showing both the batch code and a printed manufacture date. Send one.