The Lockout Report

Loofae is a small startup. We build a daylight-powered backup power accessory for smart locks, so yes, we have a stake in this topic — and we would rather you know that upfront.

When we looked for a credible answer to “how often do smart locks actually lock people out?”, we found marketing claims and anecdotes, but no transparent, checkable numbers. So we built one. The conclusions below stand or fall on the data and stated assumptions.

We examined Amazon US customer reviews exported in April 2026 for 14 smart lock listings spanning leading marketplace brands. The sales-matched set contains 12,275 reviews; 68% were 5-star, 10% 4-star, 5% 3-star, 4% 2-star, and 13% 1-star.

Samples from every category were manually spot-checked. Keyword matching can miss misspellings, sarcasm, or unusual phrasing; we estimate classification error at ±0.3 percentage points. Reviews show what buyers complain about, not a random sample of all owners, so we separate documented counts from estimates.

Matching the review and sales windows gives 107 documented lockout reports in 2025 — roughly one report for every 8,000 locks sold. These are buyers who were locked out and then returned, often months later, to write about it.

The true rate cannot be observed directly because only a fraction of buyers write reviews. Depending on the assumed review rate, the implied annual incidence ranges from about 1 in 800 to 1 in 160 locks.

Amazon allows one review per buyer per product, while many lockouts occur 5–24 months after purchase. Accounting for victims whose review channel has effectively closed can raise the estimate further. Across the full sensitivity range, the six-year lifetime risk runs from roughly 1 in 270 to as many as 1 in 9 owners. The documented floor requires no assumption at all.

Battery problems appear in 14.1% of 1–2 star reviews, statistically tied with fingerprint issues at 14.4%. Only one-time installation and fit friction ranks higher. A further 2.1% of negative reviews describes a battery that fully died.

Advertised battery life among the sampled locks ranges from five months to one year, yet reviewers routinely report receiving a fraction of those claims. WiFi models draw especially heavy battery complaints, consistent with their always-connected radios.

Brand-level lockout mentions ranged from 0.8% to 4.5% of reviews among the larger groups. At model level, the spread was wider, but the two highest rates came from very small samples and should be treated as directional rather than conclusive.

The three highest battery-complaint rates in the sample belonged to WiFi locks: Schlage Encode at 39.3% of 28 reviews, Wyze at 24.9%, and Philips at 21.9%. The Schlage sample is small, but the overall pattern matches the higher power demand of continuously connected WiFi radios.

The pattern aligns with battery chemistry: alkaline cells lose substantial usable capacity in freezing temperatures and degrade faster in heat. July shows a smaller secondary bump. “Lockout season” is real, and it is winter.

1. Reviews are not a random sample. Unhappy buyers may be more likely to write.
2. Incidence estimates depend on an unobserved review rate. We therefore present a sensitivity range, not a single point estimate.
3. Keyword classification has limits. It can miss unusual wording or misunderstand context.
4. Cause attribution is incomplete. Many lockout reviews do not state whether the cause was battery, mechanical, or software failure.
5. Sales figures are third-party marketplace estimates, not audited shipment data.
6. Loofae sells a product in this category. We disclose that interest so readers can assess the work accordingly.