Skip to main content
Site Care — Twelve Months of Quiet Reliability hero image

A year. Zero outages. Nobody noticed. That's the point.

Infrastructure
Site Care — Twelve Months of Quiet Reliability

A year. Zero outages. Nobody noticed. That's the point.

A twelve-month Site Care retainer for a small-business marketing site — dependency upkeep, performance monitoring, content publishes, and zero unplanned downtime.

12 months
Retainer Duration
0 hours
Unplanned Downtime
0
Security Incidents
14
Dependency PRs Merged
Problem

The challenge

Small-business marketing sites have a predictable failure pattern. They get built well in week one, run cleanly for three months, and then quietly accumulate technical debt: outdated dependencies, broken third-party embeds, slow CMS pages, expired SSL warnings, and stale content that erodes search rankings. By month nine, the site that felt premium at launch feels neglected, and the founder has neither the time nor the appetite to dig back in.

Site Care is the retainer that prevents that arc. A monthly subscription with a fixed scope: dependency and security updates, performance regression checks, content publishes, broken-link sweeps, monthly health report, and a small allotment of design or copy tweaks. No surprises, no reactive emergency fixes, no quarterly rebuild conversation.

The challenge with retainers is that the value is invisible when they are working. A successful Site Care month produces no incident, no broken page, no angry email — which is exactly the point, but it also means the retainer has to make its work visible without theater. The monthly health report has to do that job.

Approach

How we built it

Onboarding: full site audit, dependency inventory, monitoring setup. Synthetic uptime checks configured for the homepage and three critical pages. Lighthouse baseline captured for performance regression detection. Backup and rollback procedure documented.

Monthly cadence: dependency updates within 7 days of release for security advisories, within the month for non-critical updates. Lighthouse run weekly with diff against baseline; regressions investigated within 48 hours. Broken-link and 404 sweeps monthly. Content publishes within 2 business days of submission.

Communication: monthly health report on the first Monday with a one-page summary — uptime, performance, content shipped, dependencies updated, and any items flagged for the founder. No status calls. No meeting overhead. Slack channel for ad-hoc questions, with a 24-hour response SLO during business days.

Architecture

System map

How the pieces talk to each other.

Site Care Retainer ArchitectureA scheduled monitor fans out across dependency, performance, uptime, and backup checks, aggregates into a dashboard, and produces a monthly report for the client.Cron Monitorhourly + dailyDep Updatesrenovate / npm auditPerformancelighthouse + RUMUptimemulti-region pingsBackupsDB + assetsSecurity ScansOWASP / SSLDashboardstatus + historySlack Alertscritical onlyEmail DigestweeklyMonthly ReportPDF + summaryrunmetricspagerolluptickDASHBOARD CYCLELIVE METRICS
Built UI

Selected screens

Real product surfaces from the engagement — not stock illustrations.

Uptime monitor showing 12 sites at 99.98% uptime over 30 days
1 / 2

Uptime status — 12 sites, 99.98% uptime over the last 30 days, no SEV pages.

Build

What shipped

Synthetic uptime monitoring on 4 critical paths (homepage, services, contact, key landing page), 1-minute interval, alerting via email and Slack.

Weekly automated Lighthouse runs with budget thresholds (LCP < 2.5s, CLS < 0.1, TBT < 200ms); regressions auto-create a tracked task.

Monthly dependency update PR with changelog summary and test results; security advisories handled within 7 days regardless of cadence.

Monthly broken-link sweep across the entire site; redirect map maintained for any moved or removed pages.

Content publishing pipeline: founder submits markdown, copy, or assets via shared folder; content lives on production within 2 business days, including image optimization and OG image generation.

Monthly health report: PDF with uptime numbers, performance trend, dependencies updated, content shipped, and a forward-looking risks-and-recommendations section.

Outcome

Results

12-month retainer completed with zero unplanned downtime and zero security incidents. All dependency updates landed within the cadence policy. Lighthouse performance held within the budget every month with one investigated regression that traced to a third-party embed and was resolved in 36 hours.

Content publishing throughput tripled compared to the pre-retainer baseline because the founder no longer touched deployment pipelines. Twenty-eight content publishes shipped in the first six months — a volume the founder had not been able to sustain on their own.

Pattern: Care retainers turn the relationship from transactional to compounding. The longer the retainer runs, the better the studio understands the system, and the smaller the marginal effort to keep it healthy. Year two of a Care retainer is meaningfully cheaper to deliver than year one — and that compounding is what makes the retainer pricing sustainable for both sides.

Artifacts

Available

  • Sample monthly health report (anonymized PDF)
  • Onboarding audit checklist
  • Dependency update policy document
  • Lighthouse budget configuration
  • Content publishing workflow document
References

Talk to people on this work.

No fabricated quotes. Reference contacts are shared during discovery, with both parties' consent.

Reference available

Engineering lead

Fintech · 5 years

Worked alongside on production trading systems for 5+ years. Available for technical reference calls — code quality, on-call discipline, incident behavior.

Reference call shared during discovery, both consenting.
Reference available

Founder

Studio engagement

Engaged Sage Ideas for a Ship + Operate combination. Willing to talk about scope discipline, timeline accuracy, and what handoff actually looked like.

Reference call shared during discovery, both consenting.
The most valuable retainer hour is the one nobody noticed. The deploy that didn't break. The dependency you didn't have to read about.
// build log · entry 04
Honesty

What almost happened.

Every project has near-misses. Decisions that, if we'd kept going, would have shipped a hole. The list below is the diff between the version that almost made it to prod and the version that did.

// near-miss · 01
diff
-
beforeRenovate was going to merge dependency PRs automatically, including a major Next.js bump that would have shipped on a Friday afternoon.
+
afterAuto-merge gated by suite color + a freeze window that closes for the weekend. Majors require a human ack regardless of CI.
$
costSlightly slower upgrade cadence. Zero weekend rollbacks in 12 months.
livebuild 29be8ec2026-06-11 06:38Z
// solo studio// no analytics resold// every commit human-reviewed