Compatibility, network conditions, install and upgrade paths, battery behaviour. Run on the actual models sitting in your crash logs, by paid testers who own them.
Smoke and flow passes across the device list you supply, in the order of your crash volume.
2G, flaky café Wi-Fi, captive portals, mid-session handover to 4G and back.
Fresh install, in-place upgrade, restore from backup, and cross-version migration.
Background drain and heat measured over real usage windows, not a benchmark loop.
Repeat suites distributed to testers who own the exact hardware, on your release cadence.
VoiceOver and TalkBack passes, hit-target checks and dynamic type.
You send the build, the device list and the flow that matters. We send a scope, a date and a fixed fee before anyone installs anything. Reports are video plus logs, ticket-ready, and every finding is retested free inside 14 days.
We are small on purpose: one studio lead, a standing panel, no account managers in between. Nothing gets subcontracted to a lab you cannot ring.
Redacted, because that is the deal. Device counts and turnaround are real.
38 handsets, A9–A14. Two 3DS handoff defects on A9 and A11, plus a keyboard overlay on the CVV field.
4 days · 3 blockers22 devices. Date picker unusable with VoiceOver on iOS 15, and a crash on second guest entry.
2 days · 2 blockers54 devices. Flaky slot picker on 3G, and an upgrade path that lost the saved basket on 3 of 12 models.
5 days · 6 findingsUpdate install loop under low storage. Found on 4 of 30 units, all below 2GB free.
3 days · 1 blockerBuild, flows, device list, deadline. If the list is vague we help build it from your crash reports.
Fixed fee, dates, session count, retest window. Signed before anyone installs anything.
Briefed testers on their own hardware, video plus logs, uploaded as they go rather than at the end.
Every blocker retested free inside 14 days, closure recorded per defect.
The handsets everyone else threw away are the ones that find the bugs. If you have one in a drawer and 20 spare minutes, you can be paid to run a brief on it.
No. A signed build list and install links are enough. Repo access only helps if you want build-level notes.
Two days for a focused sweep, five for a wide matrix. Same-week starts are usually possible.
For breadth, yes, but anything the report leans on comes from hardware. Emulators miss thermal and radio behaviour.
Only testers on that run, under a written confidentiality note. Access to build links is revoked at the end.
Tell us the app, the platforms and the deadline. Reply within a working day, scope attached. If your stack is out of scope we will say so rather than pad.
THE QUIRKY QA LTD · 128 City Road, London EC1V 2NX · Company No. 13091946