All practicesRun
Managed Services — Essential
Their live platform is monitored around the clock, with BigStep engineers responding to outages 24×7 and to other issues in business hours, under written response times, plus a monthly report on uptime, incidents, cloud spend and changes.
Managed Services — Essential
$5K–$10K a month
- Duration: Monthly
- Tier: Run
- Format: Monthly
You leave withMonitoring, alerting, incident response and a monthly health report
Book a Free SessionSound familiar?
Our platform is live and brings in business, but we don’t have anyone on call. If it goes down at night, we find out in the morning from a customer. The team that built it has moved on, and there is nobody left who knows the system well enough to fix it fast.
What you leave with
Uptime checks and dashboards
Format: Client’s cloud monitoring (GCP Cloud Monitoring / CloudWatch / Azure Monitor)
How we know it’s acceptableEvery covered service has a check, the client can view the dashboard
Alert rules and on-call routing
Format: Terraform in the client’s repo
How we know it’s acceptableA test alert reaches the on-call engineer within the P1 response time
Runbook
Format: Markdown in the client’s repo
How we know it’s acceptableAn engineer new to the system can restart or roll back using only the runbook
Escalation matrix
Format: Document
How we know it’s acceptableNames, 24×7 contacts and response times confirmed by the client
Monthly health report
Format: PDF, 1–2 pages
How we know it’s acceptableDelivered by the agreed date each month; covers uptime, incidents, response times, spend and changes
Also included
- Month 1 onboarding: access, service map, monitoring set-up, runbook, baseline.
- Written response times: P1 (service down or major feature broken): response within 1 hour, 24×7. P2 (degraded, partial failure): within 4 business hours. P3 (minor issue, question): next business day.
- Fixes needed to restore service after an incident, plus a short root-cause note for every P1.
- Monthly checks with early warnings: SSL certificate expiry, WAF status, cloud budget alerts, failed backups.
How it runs
Monthly
Onboarding
Week 1
Access and discovery
Get access, map services and integrations, review Terraform and CI/CD.
OutService map and a list of gaps.
Weeks 2–3
Set-up
Uptime checks, alert routing, P1/P2/P3 definitions, runbook, baseline uptime and spend, escalation matrix.
OutMonitoring live, runbook, response times switched on.
Then, on a repeating cadence
Every month
Run
Monitor, respond, restore, run the monthly checks.
OutMonthly health report.
Every quarter
Review incident patterns.
OutA short note recommending staying on Essential or moving to Growth.
Is this right for you?
This is for you if
- Buyerthe person accountable for the platform after launch, usually a CTO, Head of Engineering, Head of Digital or technical founder at a small-to-mid-sized company without its own on-call team.
- Also has to say yesthe client’s IT/security owner (approves cloud access) and Finance, because this is a recurring monthly cost.
This is the wrong answer if
- The system is not live yet. Monitoring a prototype has no value; the client needs a Build sprint (the AI Feature Sprint, the Agent Prototype Sprint) first.
- The client expects regular feature work or bug fixing every month. That is a development retainer or Engineering Pod, not Essential.
- The client wants their cloud bill reduced, pages made faster, or the AI agent improved. That is Growth or Inference Cost Teardown.
- The client needs a guaranteed uptime percentage with financial penalties, or 24×7 response to every issue, not just outages. That is a custom contract priced well above Essential.
- The client will not give us cloud access, logs and rights to restart and roll back. We cannot respond to what we cannot see or touch.
- Infrastructure is undocumented and not in code. A paid hardening step is needed before we take on the service.
What’s included, and what isn’t
In scope
- 24×7 monitoring and alerting
- P1 response 24×7
- P2/P3 response in business hours
- Fixes needed to restore service
- Root-cause notes for P1
- Monthly checks of SSL, WAF status, budget alerts and backups
- Monitoring config in Terraform
- monthly health report
- Runbook upkeep
- One production environment (plus staging at the top of the band).
Out of scope
- New features and UI changes
- Bug fixes not tied to an incident
- Cost optimisation, performance tuning and load testing (Growth)
- AI agent evaluation or model upgrades (Growth)
- Analytics, lead or SEO reporting
- Content and data updates
- Security changes beyond restoring service (for example, new WAF rules)
- Re-architecture
- Cloud bills and third-party licences (paid by the client)
- Uptime guarantees or service credits.
Before we start
Before day one
- Cloud console access with agreed IAM roles, including permission to restart services and roll back deployments.
- Access to the code repository, CI/CD pipeline and Terraform state.
- Access to integrations the platform depends on, or a named vendor contact.
- A named client contact for P1 escalation, reachable 24×7.
- Signed-off list of covered services.
- Response times start only after onboarding is complete.
What we’ll need from your team
- Month 1: a technical contact for about 4 hours a week (access, walkthroughs, approvals).
- After that: about 1 hour a month to review the report.
- A decision-maker reachable 24×7 during a P1, for example to approve switching off a failing integration to bring the site back.
Pricing
What we need to give you a quote
- Cloud provider and account structure
- List of production services and integrations.
- Current monitoring tools.
- Whether infrastructure is in Terraform/IaC.
- CI/CD setup.
- Incident history for the last three months.
- Current monthly cloud spend.
- Any compliance or data-residency requirements.
After we hand over
Who owns it after we leave
While the contract runs, BigStep operates it and the client’s named owner receives the report and approves changes. If the contract ends, monitoring, alert config and runbooks are already in the client’s own accounts and repo, and we run one handover session with whoever takes over.
What we measure
- Uptime % per service (measured, not guaranteed)
- P1/P2 incidents per month
- Actual response times against the committed times
- Time to restore service
- Monthly cloud spend
Baselines are captured in month 1.
What comes next
- Growthwhen the client wants cost reduction, faster pages or AI improvements. Our anchor client already receives this kind of work, which shows the upgrade path is real.
- Also Inference Cost Teardown if AI or cloud spend is high, and a development retainer if the client wants regular features.
Questions buyers ask us
Monitoring tells you something broke; at 2 a.m. someone still has to fix it. That’s what you pay for.
We guarantee how fast we respond, which is what we control. We measure and report uptime every month, so you can see the result.
Yes, as a change request at an agreed rate, so support and new work don’t compete for the same hours.
Yes. Outages get a 1-hour response day or night; smaller issues get business-hours response. Full 24×7 on everything is available as a custom plan.
Fair point. That’s why month 1 is onboarding and the response times start only once onboarding is done.