Failing Projects
Before You Spend More on a Project That Isn't Delivering
Four checks before you approve more training, another tool, or a consultant. You can run them without a vendor.
Executive Summary Check the fit before funding the next fix. +
The project has slipped, the launched system gets bypassed, or the last fix faded. Another round of spending is on the table.
Run four checks first. Size the business consequence. Confirm the basic standards are clear, workable, and followed. Name what an outside provider would add that your team cannot. Decide now how the gain holds after the extra attention ends.
The right answer may be an internal repair, a different specialist, or stopping the project.
The Stalled Priority Snapshot is a 90 minute working session on one project, with a written report within 24 hours. Recommendations follow the evidence. The fee is $1,500.
You have already put time, money, and credibility into this project. The deadline moved. The system launched, and people still route around it. The last fix held for three weeks, then the old pattern came back.
Now someone wants the next intervention approved: more training, another tool, extra staff, a consultant, a larger change program. You have approved versions of this before. Before you sign this one, you need a reason to believe it fits the problem. Four questions supply one.
What business consequence would justify the next step?
Imagine a customer onboarding project that keeps missing its promised dates. A customer who signed in March is still waiting in May. Two coordinators spend their afternoons chasing approvals and re-entering data. A vice president is personally walking accounts through the queue.
Start with what was missed, where the work waited, what had to be redone, and what it cost customers. Establish the result that matters, the size and frequency of the shortfall, and what happens if it continues. Confirm the result is still worth pursuing.
Keep different consequences separate. Rework consumes time. A delayed payment affects timing. A lost customer affects future business. Each needs its own evidence, and counting the same consequence twice inflates the case for action.
Then weigh the likely value of the next step against its fee, internal effort, and implementation burden. A short investigation can be worth it if it resolves uncertainty before an expensive decision. A larger intervention needs a stronger case.
Money already spent explains why the situation matters to people. Justify the next commitment by what it can still accomplish.
Are the basic standards clear, workable, and followed?
Paul Stansik's "On Execution" asks whether the basic work is being done consistently, since standards erode as inspection, training, and reinforcement fade.
Apply that to the work itself. In the onboarding example: is there an agreed definition of a complete handoff? Does someone own the approval? Is the expected response time clear? Can the team detect a missed commitment and act on it?
A workable standard that has drifted needs an owner, a reminder of what it is, and someone checking. A capable manager can do that this month without a budget line.
Also check whether the standard can be met at all under current conditions. An approval rule that routes every decision to an executive who is never available is a constraint. So is a deadline that depends on information nobody can get. People need the skills and information to do the work, and the authority and time to act. Where those are missing, fix them before you buy anything.
What would outside help add?
Name the gap an engagement would fill. Perhaps each department has a plausible story and nobody has reconstructed the full sequence. Or an earlier fix failed and nobody can say why.
Stansik's "Count the Wolves" makes the same case: get close to the work, measure the problem, and bring in an outside view when it adds something.
Ask what the provider will establish that your team and current advisers cannot. What evidence will they use? What decision becomes possible? Who has the authority to act on the finding?
In the onboarding case, a useful contribution would be showing where the delay comes from:
- Incomplete inputs.
- An approval queue.
- Repeated changes.
- A practice people have stopped following.
Each points to a different action.
If your team can establish the cause, make the change, and check the result, an internal fix is enough. Spend the money elsewhere.
What would make the improvement last?
Before the change starts, agree how it becomes ordinary work. In "What Does an Operating Partner Do?", Stansik ties standards to agreement, measurement, and hands-on support through implementation. For your intervention, specify:
- What will change and who owns it.
- The starting performance and the result you expect.
- The support people need to make the change.
- When you will review progress and what would trigger an adjustment.
- How you will check that gains hold as support decreases.
In the onboarding example, faster approvals during a week of executive attention would be encouraging. It would not tell you whether the process holds under normal demand, whether errors rose, or whether another team inherited the delay.
Record other changes that could explain the result, such as lower volume or extra staffing. Agree in advance what would justify continuing, expanding, changing, or stopping the intervention.
An ongoing service needs an ongoing job: keeping a practice alive, helping a connected team adopt it, catching drift before it costs anything. If you cannot see that job, stop paying for it.
Where Emergent Skills fits
Run these questions on Emergent Skills too.
ES traces how work moves and where it waits, loops, or comes back for rework, and weighs the other explanations alongside: unclear priorities, missing resources, dependence on an outside party, inconsistent execution. Where the evidence points there, it also checks whether the load on people is getting in the way of work they already know how to do.
The Stalled Priority Snapshot covers one result and the decision about what to do next. If the evidence shows a recurring pattern, the Work Demand Diagnostic workshop examines it before anyone commits to deeper measurement. Anything after that has to pass the test above: an improved result, evidence about its cause, and a way to keep it.
Put the decision on one page
Before approving another intervention, write down:
- The result that is falling short.
- Its business consequence.
- What you have already tried.
- What remains uncertain.
- The decision you need to make next.
Then ask the proposed owner or provider how the next step helps you make that decision or improve that result, what it will need from your team, and how you will judge its value.
That page is where a conversation with ES starts. It is also how you decide whether you need one.
Bring one project that is not delivering.
Ninety minutes on one project. A written report inside 24 hours. The fee is $1,500. It may point to the workflow, the priorities, or the resources. If the right move is an internal repair, the report says so.
Projects stall, results fall short, and problems keep coming back. Trace what happened, test the explanations against the evidence, and decide what’s worth changing. 
