How cookware dies: 1,186,236 products, a 2-year median, and a volume effect
Across 1,186,236 products, the median review lifespan is 2 years and 46.9% are discontinued. The strongest predictor of ending is not type — it is how many reviews a product already has.
What “ended” means
A product’s review curve is ended when its last sampled review is at least three years before the 2023 snapshot. That is a statement about the review record, not about the Amazon listing: a product can stop collecting reviews and still be on sale, or be delisted while reviews continue for a while.
Across 1,186,236 products, the median review lifespan is 2 years, and 46.9% are classified as discontinued.
Type explains less than volume
The ended share varies by product type, but not by much: from 67% for cutlery sets and tagines to 81% for roasting pans, across the 30 types we publish. The type is a weak explanation.
The strong pattern is review volume:
| Minimum reviews | Products | Share whose review curve ended |
|---|---|---|
| 1+ | 1,186,236 | 63% |
| 10+ | 256,490 | 43% |
| 50+ | 59,150 | 29% |
| 100+ | 28,792 | 22% |
| 300+ | 7,748 | 11% |
| 1,000+ | 1,400 | 3% |
| 5,000+ | 68 | 0% |
A product with a long review history is unlikely to simply vanish from the record. A product with one review is mostly noise: it can disappear without telling us anything about quality.
What it means for a buyer
- Do not read disappearance as a quality verdict. A product can end because the seller stopped it, the model changed, or the listing moved.
- Use review volume as a stability signal. A product with thousands of reviews has a track record; a product with a handful does not, regardless of its star rating.
- Check the listing status separately. The lifecycle column separates the review curve from the live Amazon check; a delisted product did not always stop selling, and a product with an ended curve may still be on sale.
- For a specific purchase, the product page matters more than the type average. The type-level ended share is a base rate, not a forecast.
What this does not tell you
“Ended” is defined from the review curve, not from sales, inventory or a delisting event. The 2023 cutoff makes recent years partial. Products with no reviews at all are not in the sample. The volume table is descriptive: it shows that high-volume products are less likely to disappear from the review record, not that reviews cause longevity.
Evidence behind this investigation
- Lifecycle columnPhases, median lifespan, ended share by type, live listing status
- Brand lifecyclesPortfolio trajectories for 120 brands
This investigation is review-derived (2023 vintage). Rates are the share of sampled reviews that mention a problem, not a lab-measured failure rate; associations are not causation. See the methodology, scoring standards and editorial policy. Corrections: how to report one.