Rescuing a Stalled Project: Three Moves

A stalled project is rarely the one on fire. It is the one that has been ninety per cent complete for nine weeks, where the weekly status is still green, the team is still working, and the end date has quietly stopped moving. Nobody is hiding anything. The project has simply lost one of the three things it needs in order to move, and the usual response, cutting scope, repairs the one that was never broken.

The short answer. Rescue a stalled project with three moves, in this order. Freeze delivery for about three days and establish the real position. Name one person who may decide what happens next, and write down what they may decide without asking again. Then cut the scope to a core you can finish with the capacity and the weeks you actually have. The order is not a preference. A decision taken on a fictional position is a guess, and a cut made without a decision owner is reopened within a month.

A working definition. A stalled project is one where the work continues but the completion date stops moving. It is not a late project. A late project has a new date that somebody owns. A stalled project has an old date that nobody believes and nobody changes.

What does it mean to say a project has stalled?

A project has stalled when its reported position and its real position have come apart, and no mechanism is left to close the gap.

The tell is not slippage. It is flatness. The percentage complete is identical for three reporting cycles in a row, the same two risks are carried forward in the same words, and nobody has asked for a new date. Lateness is a reporting event, and somebody has to write it down. A stall is a reporting failure, so nothing gets written down at all. The anatomy of why projects stall is a separate question from how you restart one.

There is also an asymmetry worth naming. Reported progress can rise without any deliverable being accepted, because a percentage is an opinion, whereas a finished deliverable is a fact somebody else has to agree to. That is why stalls cluster in the last tenth of a plan. The cheap work of claiming progress continues long after the expensive work of finishing it has stopped.

Missing a date is ordinary. In PMI's Pulse of the Profession 2021, a survey of 3,950 project professionals fielded in October and November 2020, 55 per cent of projects were completed on time, 62 per cent within budget, and 34 per cent experienced scope creep. Roughly half of all projects arrive late, which means lateness on its own tells you very little.

What is not ordinary is confidence falling while the date holds. Australia's federal assurance regime publishes this directly. In the Digital Transformation Agency's Major Digital Projects Report 2026, ten of the 52 continuing Tier 1 and Tier 2 projects that reported delivery confidence in March 2025 had dropped from high or medium high confidence to medium or lower a year later, and one third of the projects now rated medium or lower are due to complete within the next six months. Those are large government programmes rather than SMEs, and we do not deliver IT implementation work, so take the shape rather than the rate: confidence collapses well before the date does, and a project whose date sits inside six months while its confidence sits at medium or lower has already stalled.

What is the stall actually costing you?

More than the delay and less than you would expect, which is precisely why it survives. The visible cost is the team's time after the date stopped moving, and in a business of fifty to a hundred people that figure is usually too small to force an escalation and too large to leave alone.

Here is the arithmetic, on an illustrative Melbourne equipment services firm. This is a modelled example, not a client. The stated assumptions: 68 people, USD 12M revenue, a ten per cent operating margin so USD 1.2M of operating profit, a customer onboarding redesign planned at twenty weeks and now in week 31, six people assigned at sixty per cent each so 3.6 full time equivalents, a fully loaded cost of USD 480 a day, and nine weeks reported at ninety per cent complete.

The stall itself. 3.6 equivalents across five days for nine weeks is 162 person days. At USD 480 that is about USD 78,000 of capacity spent after the date stopped moving, or 6.5 per cent of a year's operating profit.

Now the sanity check, because that number has to sit against total capacity. With roughly forty people on delivery across 220 working days, the firm has about 8,800 person days a year. The stall consumed 1.8 per cent of it. That is the whole reason nobody escalated. A stall is never large enough to look like an emergency on a capacity line.

And here is the number worth carrying. A three day freeze with all six people in the room costs 18 person days. One more week of the stall also costs 18 person days, because 3.6 equivalents across five days is the same figure. The freeze is not an expensive intervention. It is a week you were going to spend anyway, spent on purpose.

We have deliberately not priced the delayed onboarding revenue, which in most cases dwarfs the capacity number. The value of finishing is specific to the business, and a figure we invented for it would turn this model into a sales calculator. Two honest directions of error: at a lower allocation than sixty per cent the number falls, and if the stall has pulled in people outside the team to produce status and reassurance, which it almost always has, it rises.

Why do most project rescues fail?

Because they begin with the cut. Of the three moves, cutting scope is the only one that produces an artefact you can put in front of a board, so it goes first, and it goes first on a position nobody has verified and with nobody holding the decision.

We call the alternative the Rescue Sequence, and the sequence is the whole of it. A project moves forward only when three conditions hold at once. Somebody knows where it actually is. Somebody is allowed to decide what happens next. What remains fits the capacity and the weeks that are left. Remove any one of the three and the date stops moving. There is no fourth condition, because a project where the position is known, the decision is owned and the remaining scope fits is not a stalled project at all. It is a plan with a date.

The order is forced rather than chosen. You cannot decide on a position you do not have, so the freeze precedes the decision. A cut agreed without a named decision owner is reopened by whoever was not in the room, so the decision precedes the cut. Run the three moves in any other order and you produce a second stall, usually inside a month, and the second one is harder because the team has now watched a rescue fail.

Exhibit 1: The Rescue Sequence

Move	What it establishes	The test it must pass	What skipping it costs
1. Freeze	Where the project actually is, counted in deliverables finished rather than percentages reported	A written position that the delivery team and the sponsor both sign	You replan from fiction, and the new date fails in the same way as the old one
2. Decide	Who may change scope, date or budget, by how much, without asking again	One name, one written mandate, one stated limit	The cut is reopened by whoever was not in the room, and the stall restarts
3. Cut	A core that fits the capacity and the weeks you have left	A deliverable a customer would notice, finishable in weeks rather than quarters	You keep the whole scope and simply add the whole delay

Move 2 is the one almost everybody skips, and it is skipped for a structural reason rather than a careless one. The freeze produces a document. The cut produces a smaller plan. Move 2 produces nothing you can show anyone, which is why it falls out of a rescue that is already under pressure to look decisive. It is also the move that holds the other two in place, because it settles who owns the date.

What do the three moves involve in practice?

Three days, one page and one meeting, in that order, and all three inside a fortnight.

Move 1, the freeze. Stop delivery work, not the project. Count what is genuinely finished and accepted rather than what is reported as a percentage, and ask every person on the team the single question that surfaces a stall: what are you waiting for, and from whom. What makes this work is that it is not run by whoever wrote last week's status. That independence is exactly the thing being quietly dropped. In the 2021 KPMG Australia and AIPM Project Management Survey, published in November 2021 with 473 Australian project practitioners, only 49 per cent said their organisation undertakes independent project performance reviews, down from 59 per cent the year before.

Move 2, the decision. One name, one mandate, one limit, written in a sentence anyone can quote: this person may change scope and date within these bounds, without further approval. The same Australian survey found 27 per cent of respondents saying sponsors were actively engaged in fewer than half of their projects. An absent sponsor is not a governance irritation. It is a stall generator, because every decision that needs them queues, and a queue of decisions is what a stall is made of.

Move 3, the cut. Reduce the scope to a core that passes two tests: a customer would notice it, and it is finishable in weeks rather than quarters. If no such core exists, the honest output of the sequence is to stop the project, and that is a result rather than a failure.

What should I do on Monday?

Book the freeze before you reopen the scope. Five steps, and the first one takes twenty minutes.

Pull the last eight status reports and lay them side by side. Mark what actually changed between them. If the percentage complete and the top risks are the same for three cycles running, you are not late, you have stalled, and the rest of this list applies.

Put three days in the diary this week for everyone on the delivery team, and tell them plainly that delivery stops for those three days. Appoint somebody who did not write the status to run it, which in a small business usually means somebody from outside the project.

Before the freeze ends, write the decision mandate: one name, one limit, one sentence, and put it on the same one page project plan that carries the date. Only then open the scope conversation.

Tell the customer or the internal stakeholder on the Monday rather than at the end of the fortnight. A three day pause that you announce reads as control, and the same three days discovered afterwards reads as a project in trouble. This matters more than it sounds, because the fear of that conversation is the single most common reason a stall is allowed to run for another month.

What you will not have by Friday is a new date, and it is worth saying so out loud before you start. The position is the output of the freeze. The date is the output of the decision, and a date produced before the decision is just the old date wearing a different number.

Frequently asked questions

How long does a project rescue take?

The sequence itself runs inside a fortnight: about three days for the freeze, a few days to settle the decision mandate, then one session to agree the cut. What takes longer is delivering the reduced core, and that is a function of what you cut to rather than of the rescue. If a rescue is still running after a month, it has become the project.

What does it cost to bring in an outside operator to run the rescue?

Rem.Up Fractional Project Managers start from USD 3,000 a month for one day a week, billed as days worked against a day rate, and a rescue typically needs more days in the first fortnight and fewer afterwards. Pricing is a monthly retainer or a project fee, in USD. For the illustrative firm above, the first fortnight costs a fraction of the fortnight of stall it ends.

How do we decide whether to rescue the project or stop it?

Run moves one and two first, then ask whether any core exists that a customer would notice and that is finishable in weeks. If the answer is no, stopping is the correct output. The reason this question feels impossible before the freeze is that you are being asked to judge a position nobody has established, and the reason it feels impossible without a decision mandate is that nobody present is allowed to say yes.

If a project of yours has not moved its date in three reporting cycles, the freeze is the cheapest thing you can do this month. Rem.Up places senior independent operators as Fractional COOs and Fractional Project Managers in founder-led SMEs across APAC and Australia. Our Fractional Project Manager service puts a senior operator inside the project to run the sequence and then deliver the core, and you can book a thirty minute call to work out whether that is what your project needs.

Innovate. Optimize. Grow.