About this article: This article is reviewed against the cited sources. This is source research, not lab validation. Check the stated versions and your own requirements before making changes.
Direct answer: Service acceptance on day one should record that agreed scope, access paths, backups or exports, support channels, roles and first integration baselines are in place—not merely that VMs exist. A signed acceptance record with named owners gives both sponsor and provider a defensible start line for hypercare (elevated support immediately after go-live), billing disputes and future transition.
This is buyer↔host onboarding acceptance, not ISV tenant provisioning (Onboarding a new customer onto your hosted application platform) or post-migration decommission gates (Acceptance criteria before decommissioning the old system).
Why it matters
Illustrative scenario: new customers sometimes “go live” when remote access works once, then discover weeks later that backups were not configured for one volume, finance lacks portal access, or webhook allow lists still point at old office egress. Acceptance reduces that drift by naming owners and evidence up front.
Acceptance aligns shared responsibility: the hoster operates agreed platform layers; the customer owns business data classification, user access policy and external integration partners.
How it works
Microsoft’s shared responsibility model for cloud workloads explains that security and operations duties split between provider and customer depending on service type—the same discipline applies to managed hosting contracts even when the stack is not Azure. Map your written scope to that split before signing acceptance.
Suggested acceptance checklist (editorial design—adapt to contract):
| Area | Acceptance criterion | Customer owner | Provider owner |
|---|---|---|---|
| Scope | Named servers/apps in service | Sponsor | Service manager |
| Access | Staff reach app via agreed path | IT admin | Support |
| Identity | Admin roles assigned (not shared) | Customer admin | — |
| Support | Portal/ticket path tested | Service manager | Support lead |
| Backups / exports | Policy documented; first run logged and agreed recoverability evidence (see below) | Data owner | Operations |
| Monitoring | Alerts reach named contacts (as contracted) | IT admin | Operations |
| DNS / TLS | Public names resolve; certs valid | IT admin | As contracted |
| Integrations | Inventory filed; critical path tested | Integration owner | As assisted |
| Documentation | Living documentation pack started | IT lead | Provides configs as agreed |
| Commercial | Invoice contacts; payer role | Finance | Account manager |
Evidence to attach: screenshot or log references (not security-sensitive secrets), ticket IDs for test requests, date of first successful user login, manifest of customer-held copy if applicable.
Recoverability (do not conflate with first backup log): a successful first backup run, job log or off-site manifest proves delivery, not that you can restore business content. Acceptance should include either agreed representative restore or content-validation evidence (scoped to contract—sample file restore, database check, or documented recovery test), or an explicit open gap on the record with owner and due date. For test design concepts, see Verify a backup delivered to S3 or SFTP and Keeping a usable copy of your data when you change hosting provider—this checklist does not duplicate full recovery procedures.
Explicit exclusions: list items out of scope (e.g. desktop support at branch offices, LOB vendor tier-3) to prevent assumption drift.
Revisit acceptance when material scope is added—new application, new site, DR option—not only at initial onboarding.
What to choose and why
Use one short signed document plus a living documentation pack for the hosted estate. Sponsors should not accept on verbal assurances alone.
Pair acceptance the same week with portal versus ticket rules and admin/finance/helpdesk roles.
If a criterion fails, record conditional acceptance: a named owner explicitly accepts the specific residual risk, remediation ticket and deadline. Never automatically waive critical security or data-recovery criteria—those remain open until evidenced or escalated.
Read next
Related articles you may want next.
- HostingOnboarding a new customer onto your hosted application platformSet practical acceptance gates for tenant identity, data import, integrations and production access before a new hosted application customer goes live.Read article
- HostingAcceptance criteria before decommissioning the old system after migrationWhat evidence the migration board needs before the predecessor environment is powered down — distinct from RPO/RTO proof drills and from cutover-night technical gates.Read article
- HostingKeeping a usable copy of your data when you change hosting providerKeep a usable copy of business data when changing hosting providers. Check formats, encryption keys, application compatibility and recovery in a safe test environment.Read article
Risks and limits
Checklist is buyer education; contractual terms prevail.
Does not define technical test procedures for backups or migration cutover gates.
No acceptance ceremony was observed in lab (untested).
Sources
- Microsoft Learn — Shared responsibility in the cloud — customer vs provider responsibility framing.
- Editorial — acceptance checklist, recoverability gap rule and conditional acceptance pattern.
- Verify a backup delivered to S3 or SFTP — delivery vs restore framing (catalogue pointer).
- Keeping a usable copy when you change hosting provider — usable-copy test concepts (catalogue pointer).
Learning notes
- Acceptance is a signed scope baseline—not “we can ping the server.”
- Name owners on both sides for each criterion.
- Conditional acceptance needs a named risk owner and deadline—not silent waiver of recovery or security.
- Re-accept when scope materially grows.
- Shared responsibility should appear in the record, not only marketing pages.
This article was originally published on .