Writing, not content.
We publish when we have something worth saying about product health, incidents, feedback, or the economics of running a monitoring program. No filler, no SEO farms.
Beyond the green check: what a meaningful uptime monitor validates
A 200 OK is the weakest signal an uptime monitor can give you. Here's a practical guide to monitors that actually tell you whether your product is working.
9 min readIncident communication is a product feature: a status page playbook
A structured guide to running the status page during an incident: what to say when, how often to update, and why ambiguous language costs more than the outage itself.
8 min readNPS that means something: tying feedback to what you actually shipped
Generic NPS is a gauge chart on a dashboard. Useful NPS tells you whether the release you cut last Tuesday made users happier or sadder — and at what confidence.
8 min readRunning a digital agency without drowning in ten tool logins
Agencies maintain product health for clients using tools built for single-product teams. The result is ten logins per engineer and an operational tax that scales linearly. Here's a better shape.
7 min readThe 3am page that shouldn't have happened: tuning alert thresholds
Alert fatigue is the single most common reason teams ignore real incidents. A practical guide to threshold design, quorum, and severity routing that respects on-call sleep.
7 min readGDPR-ready by default: what the label actually means
"GDPR-ready" is on every vendor's website. Here's a practical checklist for what the label should cover — and what it usually doesn't.
6 min readHow we dogfood Brily: what broke, what we learned
We run Brily on Brily. Here's an honest account of what the first month of dogfooding found — the monitors we didn't have, the alerts that flapped, the status page we weren't ready to publish.
7 min read