What 2,853 support tickets describing an actual defect say about each product in the line: 3.0, Disc, Box, plus Keychain and Birdie+ for scale. The next slide is the one-slide summary; everything after it is the detail behind it, broken out per product.
Reviewing the Box on/off-switch design with the manufacturer this month (it's 63% of Box's own defect tickets and the single most concrete fix in the data). Reporting back next meeting.
Last quarter = May to Jul 2026. Prior quarter = Feb to Apr 2026. Rate covers 3.0, Disc, and Box only, the three products currently sold โ see the next slide for why Keychain and Birdie+ are tracked separately.
Joined to real monthly units-sold from a Shopify export, so those three are real per-1,000-units rates, not just counts. Keychain and Birdie+ sold zero units in either window (discontinued lines) but still generate tickets: a support-tail finding, not a rate. Everything here is CS tickets only, so treat every number as a floor, not a ceiling.
Keychain leads on raw count because it's the oldest, highest-volume product, not evidence it's the least reliable one.
Recent = May to Jul 2026. Prior = Feb to Apr 2026. Same months for both units and tickets.
There's no unit sold to attach a rate to, so this is pure post-sale support cost on products no longer sold, not a quality problem. Worth a deliberate policy decision on how long support continues for a discontinued unit.
3.0, Disc, Box, and Keychain each show device-detaches-and-gets-lost as a leading complaint: a carabiner, a screw ring, a run-belt clip, a door mount. Different hardware, same symptom, across the whole line. Worth one cross-product engineering review rather than four separate ones.
Full detail, all quotes, and every caveat: see the companion written report, DEFECTS-BY-PRODUCT.md.