The five warning signs a transformation is about to stall

Projects rarely move from healthy to failed in a single meeting. More often, ambiguity, delayed decisions and weak evidence accumulate while each instance remains individually explainable.

Every programme also has ordinary friction. A late interview, a difficult integration or a revised milestone does not establish that the work will fail. The purpose of warning signs is to prompt investigation while the team can still change course, rather than to turn a complex programme into a traffic-light performance.

Use the five patterns below as a structured conversation. Look for duration, recurrence, consequence and the quality of the response. One serious control or service risk may demand more attention than several mild symptoms.

1. Stakeholders describe different versions of success

Ask the sponsor, a delivery lead and an affected service owner what a successful outcome would change. Differences can be productive. Trouble appears when the project is spending against incompatible expectations without a forum for resolving them.

“Launch the website”, “increase qualified enquiries” and “leave the old platform before renewal” are three different kinds of success: an output, a commercial outcome and a constraint. A sound programme may need all three. It should show their relationship and any trade-off between them.

Review:

• the problem the investment was approved to address;
• the outcomes and affected groups;
• baseline evidence;
• outputs required to enable those outcomes;
• constraints such as dates, obligations and capacity;
• owners for measuring results; and
• conditions that would make the investment no longer worthwhile.

Avoid forcing every outcome into a number when the measure would be artificial. “Clients can complete onboarding with less avoidable effort” can be tested through completion, support, error and research evidence. It is stronger than an invented target chosen because a board paper expects one.

Intervene when the differences are changing scope, priorities or design without a decision. Reopen the business case and agree which outcomes govern the next phase.

2. Governance discusses the same issues without deciding

Read the last three sets of steering records. Which decisions were made, who owned them, what evidence informed them and which issues returned unchanged?

A repeated agenda item can be legitimate if the team is gathering agreed evidence. It becomes a warning when “further discussion” replaces a decision, no person has authority, or delivery is forced to progress across incompatible scenarios.

Classify the decisions the programme needs:

• strategic outcome and investment;
• scope and priority;
• service or design direction;
• technical approach;
• risk acceptance;
• release readiness; and
• operational ownership.

Assign authority at the appropriate level. A steering group can provide oversight, resolve cross-organisational conflict and make material investment choices. It should not become the approval route for every operational decision.

Record decision deadlines and the consequence of delay. If evidence is genuinely insufficient, decide what will be gathered, by whom and when. That turns deferral into an owned action.

3. Activity grows while usable evidence stays scarce

Project teams can be extremely busy producing workshops, documents, architecture and environments while users and decision-makers see little that changes their understanding.

The answer is not always to “ship something” within an arbitrary number of weeks. A regulated migration or foundational security change may require substantial work before external release. The team should still produce inspectable evidence appropriate to the stage: a tested service slice, a realistic prototype, a migration rehearsal, a content sample, a performance result, an observed task or a resolved technical risk.

Ask:

• What has been tested rather than described?
• Who has seen or used it under realistic conditions?
• Which assumption changed as a result?
• What remains simulated or incomplete?
• What decision can now be made?

Beware the big reveal. When the first integrated experience appears late, stakeholders discover gaps after time and budget have reduced their options. Create learning slices that cross enough of the real service to expose handoffs and dependencies.

The source article included a useful self-criticism from Distinction. A client raised concern about the lack of visible progress and we treated it as stakeholder management rather than evidence about delivery. That was the wrong response. Anxiety can be poorly informed; it can also identify a gap that status reporting has concealed. Investigate the basis before managing the perception.

4. The business case stopped governing the work

Many business cases are used to secure approval and then replaced by budget and schedule reporting. Costs, scope and assumptions change while the expected value remains frozen in the original document.

At agreed decision gates, update:

• current and forecast cost;
• outcome assumptions and baseline;
• dependencies and adoption conditions;
• risk and possible harm;
• expected timing of value;
• viable alternatives; and
• evidence that would support continuation, revision or stop.

This is not an invitation to rewrite the case until it justifies sunk cost. Preserve the approved version and show every material change. If the original premise no longer survives contact with evidence, leadership needs that information before the remaining budget is committed.

An overrun may still be justified by the expected value. An on-time project may deserve to stop if the need has disappeared or the product will not be used. Governance should manage investment value as well as delivery performance.

Our companion article on what boards should know before approving a digital transformation budget covers the pre-approval questions in more detail.

5. The delivery partner reports polish and little uncertainty

Formal milestone reporting is necessary. A programme also needs a safe route for emerging technical, commercial and organisational concerns before they can be fully resolved.

Silence between milestones does not prove that a supplier is hiding problems. The contract, team structure or client preference may have created that communication pattern. It still deprives both sides of time to respond.

Distinction has experienced this from the supplier side. On one financial services project, we carried a technical risk while trying to resolve it ourselves and surfaced it later than we should have. The client had less time to act. The lesson is stronger than a generic instruction to communicate more: agree which uncertainties must be raised, at what threshold and through which route.

Create a short recurring conversation between the accountable client lead and the person responsible for delivery. Useful questions include:

• What became harder or more uncertain this week?
• Which assumption worries you most?
• What are you trying to resolve before telling us?
• Where is the client contributing to risk or delay?
• Which decision would protect the most time or value now?

This conversation should supplement the risk register and formal governance. It should not create an off-record channel where material issues remain undocumented.

Assess the pattern, not a magic score

The source article assigned points and thresholds to the five signs. That can make a conversation easier to start, though no evidence supports a universal cut-off between normal friction and imminent stall.

Instead, rate each sign using four dimensions:

Then identify interactions. Unclear success makes decisions harder. Delayed decisions prevent testable work. Weak evidence leaves the business case unchallenged. A formal supplier relationship conceals the uncertainty driving all three. The combination may be more important than the number of signs.

Escalate immediately where safety, security, legal or regulatory obligations, client harm or material financial exposure require it. Routine governance cadence should not delay a necessary control response.

Hold a recovery conversation while choices remain

Bring the sponsor, delivery and affected owners together with a one-page evidence pack. State the observed pattern without assigning motive. Confirm what has changed, which decision is blocked and what happens if the current trajectory continues.

Choose among concrete responses:

• clarify the outcome and reduce scope;
• assign decision authority;
• create an evidence-producing delivery slice;
• update or retest the business case;
• change the communication and escalation route;
• pause work that would create rework; or
• commission an independent review where internal positions have become fixed.

External review can help, and Distinction may benefit commercially from providing it. It should have a bounded question, access to evidence and independence from defending the existing plan.

The downloadable project health checklist formats these five signs for a 30-minute conversation. Use it to identify the specific intervention, not to pronounce a project healthy because it scores below a threshold.

If the programme has already stopped producing value or exhausted its authority and confidence, it needs a restart decision rather than another early intervention. The next article in this series addresses that different conversation. The important move is to recognise which situation you are actually in.