The data behind Sundr's repair-vs-replace recommendations, released free under CC-BY-SA 4.0. Attribution required. Use it in research, in articles, in training corpora, in products.
Four findings from the dataset, each with a stable link. Quote them freely with attribution.
Nokia phones are 10.5 years old, on average, when someone tries to repair one.
Apple phones average 4.1 years at repair time, Samsung 4.5. People keep Nokia phones running more than twice as long before giving up on them.
Source: 154 Nokia, 1,150 Apple, and 867 Samsung smartphone repair records (ORDS)
Philips hair dryers get fixed 82% of the time. Dyson: 47%.
Philips dryers also average 10.8 years old at repair vs 4.5 for Dyson. Price and marketing don't predict durability; serviceable design does.
Source: 221 Philips and 113 Dyson hair dryer repair records (ORDS)
Apple is the most fixable phone brand: 60% of repair attempts succeed.
Samsung sits at 46%. When an Apple phone breaks, it's meaningfully more likely to be repaired than scrapped.
Source: 1,150 Apple and 867 Samsung smartphone repair records (ORDS)
Tablets are among the hardest devices to fix.
Repair attempts succeed 47% of the time for Apple tablets and 43% for Samsung, well below the same brands' phone fix rates.
Source: 518 Apple and 513 Samsung tablet repair records (ORDS)
Curated: 206 rows where every durability score is reviewed and every fix rate has Open Repair Alliance records behind it. Full: 817 rows, 96 brands, 19 device categories, including auto-discovered models with estimated scores. Released 2026-04, curated cut added 2026-08.
https://sundr.ca/data/durability-scores-v1-curated.csv https://sundr.ca/data/durability-scores-v1.csvOne row per consumer device model. Each row pairs Sundr's editorial durability rating with the Open Repair Alliance fix rate for that device's category, plus the most common failure mode from real repair attempts.
| Column | Type | Description |
|---|---|---|
brand | string | Manufacturer name as sold. |
model_name | string | Model name as sold. |
product_type | string | Marketing device type (e.g. Smartphone, Vacuum). |
category_slug | string | Sundr canonical flat category slug (e.g. smartphone, vacuum). |
durability_score | integer 1-5 | Sundr editorial durability rating. 5 is best. |
ifixit_repairability_score | integer 1-10, nullable | iFixit repairability score where one exists. Higher is easier to repair. |
fix_rate | number 0-100, nullable | Percent of repair attempts that succeeded, from Open Repair Alliance, aggregated at the brand and category level. |
fix_rate_sample_size | integer, nullable | Count of ORDS records backing fix_rate. Small samples are indicative, not definitive. |
top_failure | string, nullable | Most common failure category for this brand and category, from ORDS. |
why_durable | string, nullable | Short editorial note on why this model scores well. Only populated for curated picks. |
The durability score is editorial, not algorithmic. It's built from multiple signals:
Editorial judgment resolves conflicts between sources. When manufacturer-reported scores disagree with real repair data, the real repair data wins. Full methodology is in the README.
Released under Creative Commons Attribution-ShareAlike 4.0. You can use it commercially. You have to credit Sundr. If you redistribute a modified version, it has to stay CC-BY-SA.
Suggested attribution
Durability scores and fix rates from Sundr (https://sundr.ca/data), licensed CC-BY-SA 4.0. Fix rates derived from the Open Repair Alliance ORDS dataset.
The canonical copy lives here on sundr.ca.
Drop a note at hi@sundr.ca. We keep a list of known uses and we're always curious what people build on top of this.