Sign-off tends to happen when the visible work looks finished. But the things that cause a bad first week are mostly invisible: undocumented patching, an untested failover, a licence assigned to a departed contractor. Here is what we confirm before calling a site complete.
Documentation
- As-built floor plan with outlet numbering that matches the physical labels
- Patch schedule mapping every port to its outlet and switch
- Certification results for every permanent link
- Rack elevation, including power distribution and UPS capacity
- Device inventory with serial numbers, warranty dates and firmware versions
Network
- VLAN plan documented, with which devices belong where and why
- DHCP scopes sized with headroom, reservations recorded
- Internet failover actually tested by pulling the primary link
- Wi-Fi validated by walking the floor, not by reading a planning tool
- Guest network genuinely isolated from internal resources
Security and access
- Default credentials changed on every device, including cameras and access controllers
- Admin accounts assigned to named people, with a documented break-glass account
- Backups running and — the part usually skipped — a restore tested
- CCTV retention verified against the required number of days
- Access control groups mapped to roles, with leavers already removed
A backup that has never been restored is a hypothesis, not a backup.
People
A short walkthrough with whoever will be first to notice a problem: reception, office management, the internal IT contact. Where the rack is, how to raise a ticket, what the escalation path looks like out of hours. Ten minutes here prevents most week-one calls.
And then a review date
Book a check-in about a month after occupancy. Real usage always reveals something the plan did not: a meeting room that is busier than expected, a printer in a dead spot, a department that grew. Fixing those while the project is fresh is far easier than raising them as new work six months later.
Related services
