When work stalls across time zones, the cause may be missing context, an unresolved decision, or a dependency on another team. Learn how to identify the constraint and choose a practical fix.

A well-written handoff can explain exactly why a task is blocked and still leave the next engineer unable to continue. The relevant discussion is linked, the risks are clear, and the person who can make the decision has finished for the day. More documentation may make the situation easier to understand without making it easier to resolve.

Before adding another meeting or asking for better updates, find out whether the engineer lacked information, needed someone else’s decision or depended on another team’s work. Each can leave a ticket blocked overnight. The useful management response depends on what actually stopped progress.

Consider a hypothetical team adding retry handling to a background job service. A request times out, but the service may already have accepted it. Retrying could create duplicate work; declining to retry could leave a legitimate request unfinished. The engineer’s handoff explains both possibilities, includes a way to reproduce the timeout and asks how the system should behave.

If the service already has a safe retry contract covering this case, the engineer may simply be missing that context. The contract needs to explain how repeated requests are handled and the limits of that behavior. Making it available with a worked example alongside the interface documentation can settle the question and give the next engineer a way to find the answer without tracking down its author.

Suppose instead that nobody has agreed how the service should behave after this timeout. Another page describing the alternatives cannot settle the choice. Someone needs to own the decision, gather the necessary product and technical input, and decide whether to accept a limitation or change the service. Naming that person helps resolve this instance. Establishing which similar decisions the team can make helps prevent the next one from joining the same queue.

Those boundaries need to be specific enough to use. An agreement might allow engineers to adjust retry timing within established limits while requiring review for changes that affect duplicate processing. A general instruction to use their judgment leaves room for an engineer to act reasonably and later discover that an undocumented approval was expected. Clear authority lets the team proceed within known limits and bring the consequential choices to the right people.

Now suppose every change to retry behavior requires the application team, the platform team and a service owner to coordinate implementation and deployment. Even a prompt decision leaves work waiting on other teams. This is where the CTO needs to distinguish a necessary joint effort from a dependency that repeatedly delays routine changes.

A change that affects several services may reasonably need a coordinated release. Plan that work together and involve the relevant reviewers early. But if routine application changes repeatedly wait for the same platform engineer to run tests, investigate what prevents the application team from running those checks itself. If the teams must release together because one service breaks compatibility with another, examine whether a compatible transition could separate the releases. The recurring wait gives you a specific engineering question to investigate before considering a wider reorganization.

DORA’s research on software delivery supports examining teams’ ability to change, test and deploy independently. Its guidance also makes clear that adopting a particular architecture does not establish that independence. For a distributed team, the practical application is to inspect the approvals, environments and service dependencies behind a delay before prescribing a new tool or structure. DORA: Loosely coupled teams

Some waits will remain justified. Security review and specialist expertise can be necessary, and scarce reviewer capacity needs to be planned. Across time zones, identify that need before the change is ready for approval and agree when the reviewer can respond. A blocker posted at the end of a working day should carry enough context for the next person to act, without requiring its author to stay online. Urgent production incidents need a separate escalation route and agreed coverage.

When the unresolved issue needs people to examine assumptions together, a live discussion may be the quickest way forward. GitLab’s remote meeting guidance combines asynchronous work with necessary meetings, supported by advance context, written records and ways for absent colleagues to contribute. Apply that to the decision at hand: involve the people needed to resolve it, allow input across time zones, and record the decision and its reasoning where the implementation team will look for it. GitLab: All-remote meetings

At the next delivery review, take one recurring delay and reconstruct the point where progress stopped. Ask the engineers involved what information, decision or completed work they were waiting for. Choose the smallest change that addresses that constraint, then follow the next comparable piece of work through review and release. Check whether the same wait recurs, whether required reviews still happen, and whether the change works as intended. That gives you evidence about delivery that message counts and hours online cannot provide.