Remote project management is often described as the same job done over video call. That definition is not useful, because it hides the one decision that determines whether the model works: which parts of the project are management and which parts are execution.
In an industrial project, execution is physical. Steel is erected, equipment is set, cable is pulled, a line is commissioned. None of that can be done remotely, and nobody serious claims otherwise. Management is a different activity: planning the sequence, coordinating the interfaces, tracking commitments, measuring progress, controlling cost and change, and telling the organisation the truth about where the project stands. That work is information work, and information work has not required physical co-location for some years now.
A working definition
Remote project management is the practice of owning the planning, coordination, control and reporting of a project from outside the site, while the client's teams, contractors and suppliers execute the physical work.
The word that matters in that sentence is owning. A remote project manager who only observes and reports is an expensive status writer. A remote project manager who owns the discipline chairs the meetings, sets the deadlines, chases the open items, escalates what is stuck and is answerable when the schedule moves.
The responsibility split
Every remote engagement that fails, fails on ambiguity about who owns what. The split has to be written down before the first meeting, and it usually looks like this.
| Remote project manager | Client organisation |
|---|---|
| Builds and maintains the plan and schedule | Provides durations, constraints and approval |
| Runs the coordination cadence | Attends and commits on behalf of their function |
| Tracks every open action to closure | Executes the actions |
| Assesses change impact on cost and time | Approves or rejects the change |
| Maintains the risk register and chases mitigation | Accepts risk and owns mitigation actions |
| Produces the reporting pack | Makes the decisions the reporting supports |
| Plans and sequences site work | Supervises and executes site work safely |
What the week actually looks like
Six activities repeat every week in a functioning remote engagement. If any of them is missing, the model degrades into reporting.
- Coordination meeting. One fixed slot, one agenda, one set of minutes with named owners and dates. Not a round of updates — a working session on what is blocked.
- Commitment follow-up. Every open action chased individually between meetings. This is the activity that separates a managed project from a monitored one.
- Schedule update. Actual dates in, forecast recalculated, critical path re-examined, look-ahead re-issued.
- Risk and issue review. New entries logged, mitigations checked, anything with a rising impact escalated in writing.
- Change screening. Anything that alters scope, cost or date is captured and assessed before it is quietly absorbed.
- Reporting. A short written status that a busy director can read in three minutes and act on.
When the model works — and when it does not
Remote project management works when the project has a competent site organisation and an information problem. It works particularly well when the client's own technical people are the bottleneck: they know exactly what needs to happen, but nobody has the hours to drive it through procurement, vendors and three contractors at once.
It does not work when what the project actually needs is a physical presence — a resident engineer, a full-time construction supervisor, a field HSE lead — and someone tries to substitute coordination for supervision. It also does not work where the client cannot or will not give access to their systems and their meetings: a remote manager with no access to the document repository and no seat in the engineering review is being set up to produce fiction.
The test is simple. If the missing capability is judgement, coordination, structure and follow-through, the model fits. If the missing capability is a pair of eyes and boots on the ground, it does not — and the honest answer is to say so.
What to ask a provider
- What exactly will you own, and what stays with us? Ask for it in writing.
- What is the reporting cadence, and what does the pack contain?
- How is progress measured, and who approves the measurement rule?
- What happens in week one, week four and month three?
- Which of our systems will you work in — and what do we keep at the end?
- What will you refuse to do because it requires physical presence?
A provider who cannot answer the last question clearly has not thought about the model carefully enough to be trusted with a capital project.