Running one IT estate across four time zones

On paper, offices in Dubai, London and Islamabad cover the working day end to end: as one team winds down, another is already online. In practice, follow-the-sun support fails at exactly one place — the seam. An incident opened at 5:40pm Dubai time, half-diagnosed, handed to London with two Slack messages and a hunch, becomes a fresh investigation instead of a continuation. The clock restarts. The customer notices.

The uncomfortable truth is that follow-the-sun is not a staffing model. It is a documentation model with staff attached. If the incident record is not written to be picked up cold — current hypothesis, evidence so far, next planned step, rollback state — then the second office is not continuing the work; it is redoing it.

The second failure mode is subtler: three offices running three variants of the same process. Different ticket fields, different monitoring thresholds, different naming conventions. Each variance is small; together they mean an engineer in one office cannot safely operate an estate configured by another. Consolidating onto one operating model — same runbooks, same tooling, same change calendar — is unglamorous work, but it is what makes the handover trustworthy.

A test we apply: could an engineer in Islamabad execute a routine change in the London-managed environment using only the runbook, with no calls? If the answer is no, the estate is not one estate. It is three estates with one logo.

Start with the incident record format — it pays back at every seam, twice a day. Then the monitoring and alert thresholds, so severity means the same thing in every office. Then the change calendar, so no office is surprised by another's maintenance window. Identity and access tooling comes next; ticketing platform consolidation, despite being the visible symbol of unification, can come last. The order matters because each step de-risks the next, and because the early steps require no procurement — only agreement.