Start by defining success
Before discussing what makes projects succeed, agree on what success means. The classic test has three parts: on time, on budget, and delivering the agreed scope. Plenty of projects pass all three and still fail. The system goes live and nobody uses it, the building is finished but never earns its expected return, or the client pays and never comes back.
So two more dimensions usually join the list: quality, meaning deliverables meet specifications and acceptance criteria, and benefit, meaning the project achieved the purpose it was launched for. For contractors and service firms there is a sixth, very practical one: margin. A project handed over on time at a loss is not a success for the company that delivered it.
What follows are ten keys that separate projects that deliver from those that drift, each with a practical tool: work breakdown structures, responsibility matrices, risk registers, change control, earned value with a full worked example, project cash flow and resource planning.
1. A clear business case and a written scope
The projects that struggle most often start with a vague sentence like "upgrade the system" or "build a new warehouse". The first step is a short document that answers four questions:
- Why? The problem or opportunity, and the expected benefit in numbers where possible.
- What? Specific deliverables, plus an explicit list of what is out of scope.
- How will we know we are done? Acceptance criteria for each deliverable.
- What are the constraints? Deadline, budget ceiling, available resources.
Scope then becomes a work breakdown structure: the project split into smaller work packages until each one can be estimated and assigned to a named owner. A warehouse build breaks into design, permits, earthworks, foundations, steel structure, MEP works, finishes and handover, and each of those breaks down further. A practical test: if you cannot estimate a package's duration and cost with reasonable confidence, it is still too big.
The out-of-scope list matters as much as the in-scope one. Many client disputes begin with an assumption that something was "obviously included".
2. A sponsor and clear governance
Every project needs a senior person who owns the decision and the outcome. The sponsor does not run the project day to day. The sponsor settles what the project manager cannot: shifting priorities, approving extra budget, resolving disputes between departments. Projects without a sponsor stall at the first disagreement because nobody can break the tie.
A responsibility matrix prevents both overlaps and gaps. For each major deliverable, record who does the work, who is accountable, who must be consulted and who must be informed. One rule matters above all: only one person is accountable for each deliverable.
| Deliverable | Project manager | Lead engineer | Finance | Sponsor |
|---|---|---|---|---|
| Baseline schedule | Accountable | Does the work | Informed | Approves, consulted |
| Budget | Does the work | Consulted | Accountable for control | Approves |
| Change requests | Accountable for assessment | Consulted | Consulted | Approves above threshold |
| Weekly status report | Accountable | Contributes | Contributes | Informed |
3. Realistic estimates and risk-based contingency
Optimism is one of the most common reasons projects slip. Better estimates come from:
- History: how long similar work actually took, not how long it was planned to take.
- The people doing the work: the engineer or technician who will deliver a package estimates it better than a manager who will not.
- Three-point estimating: optimistic, most likely and pessimistic, combined as (O + 4M + P) ÷ 6. A package estimated at 10, 14 and 24 days gives (10 + 56 + 24) ÷ 6 = 15 days, not 14.
- The real calendar: public holidays, reduced hours in Ramadan, summer heat for outdoor work, and permit lead times.
Contingency should not be an arbitrary percentage. It should come from the risk register, as described below, and be held at project level and released by decision when a risk actually materializes, not spread across packages where it quietly gets consumed.
4. An approved baseline and strict change control
Once scope, schedule and budget are approved, they form the baseline against which everything is measured. Change after that is inevitable. What separates good projects from troubled ones is how change is handled:
- Every change request is logged in writing, even when a senior manager raises it in a meeting.
- Its impact on scope, schedule, cost and risk is assessed before any decision.
- The authority named in the governance approves or rejects it.
- If approved, the baseline is formally updated. In contracting, a variation order adjusts the contract value or duration.
The most dangerous kind of change is silent creep: small additions done without logging because each looks trivial. Twenty small favors can wipe out a project's entire margin without anyone ever invoicing for them.
5. Manage risks before they become issues
A risk register is simple if it is used consistently. For each risk: a clear description, probability, financial or schedule impact, expected value, response plan and owner.
| Risk | Probability | Impact (SAR) | Expected value | Response |
|---|---|---|---|---|
| Late delivery of structural steel | 30% | 400,000 | 120,000 | Order early, qualify a backup supplier |
| Rebar price increase | 40% | 150,000 | 60,000 | Fix the price in the purchase order |
| Permit delay | 20% | 200,000 | 40,000 | Apply before detailed design starts |
| Labor shortage at peak | 25% | 120,000 | 30,000 | Pre-agree a subcontractor |
| Total | 250,000 |
Hypothetical example: the expected values add up to SAR 250,000, which is a defensible basis for contingency. As the project progresses, some risks fade or disappear and part of the reserve can be released. Review the register at every regular meeting, not just at kickoff.
6. Measure progress with numbers, not impressions
"How far along are we?" is usually answered with a percentage someone estimates on the spot. Earned value management replaces that impression with figures that compare what was planned, what was done and what was spent.
Hypothetical example: a SAR 2,400,000 project planned over 12 months. At the end of month six:
- Planned value (PV): budgeted value of work that should be done by now = SAR 1,200,000.
- Earned value (EV): budgeted value of work actually done = SAR 1,020,000.
- Actual cost (AC): what has actually been spent = SAR 1,150,000.
| Measure | Formula | Result | Meaning |
|---|---|---|---|
| Cost variance | EV less AC | SAR 130,000 over | Spent more than the work done is worth |
| Schedule variance | EV less PV | SAR 180,000 behind | Less work done than planned |
| Cost performance index | EV ÷ AC | 0.89 | Each riyal spent produces 0.89 of work |
| Schedule performance index | EV ÷ PV | 0.85 | Running at 85% of planned pace |
| Estimate at completion | Budget ÷ CPI | about SAR 2,706,000 | Roughly SAR 306,000 overrun if performance holds |
Ask this project's manager and you might hear "about half done, about half spent, all fine". The numbers disagree: the project is behind and over cost, and the projected overrun is larger than the entire contingency in the risk example. Month six is when to intervene, not month eleven.
7. Manage project cash flow
In contracting, a project can be profitable on paper and still drain the company's cash. The usual reasons: progress claims submitted late or stuck in consultant approval, retentions held until final handover, while suppliers and labor must be paid every month.
- Build a monthly project cash flow comparing expected spend with expected receipts after realistic approval and payment periods.
- Submit each progress claim on time with complete supporting documents. A missing document can push the whole cycle back a month.
- Time large purchases around expected collections where possible.
- Track retentions as an asset to be collected, not a forgotten balance.
Our guide to project budgeting and cash flow goes deeper.
8. Communication that drives decisions
- A short weekly team meeting: what was done, what is next, what is blocking. Thirty minutes at most.
- A one-page management report: cost and schedule indices, the top three risks, and decisions needed from management. Leaders need what requires their action, not every package detail.
- A decision log: date, owner and rationale for every significant decision, so it is not reopened two months later.
- One source of truth: documents, drawings and correspondence in one place, so no crew works from an outdated drawing.
9. Plan resources, do not assume them
A schedule that assumes the same engineer is fully available to three projects in the same week is not a schedule, it is a wish. Resource management means seeing the load on each key resource across all projects, not inside one.
Hypothetical example: a contractor has one electrical engineer who specializes in distribution boards, and three projects reach the electrical phase in the same month. Each plans 20 working days for him, 60 days in total for a person with about 22. At least two projects will slip. The options are to resequence work on one project, subcontract part of the scope, or delay the third start. What matters is spotting the clash two months ahead, not in the week it happens. Shared heavy equipment, cranes and testing crews need the same treatment.
10. Close properly and capture lessons
Many firms move to the next job the day after handover and lose the most valuable thing the project produced: knowledge. A proper close includes formal client acceptance and signed handover records, closing open purchase orders and subcontracts, comparing final actual cost with budget per work package, and a lessons-learned session on what worked, what to change and which estimates missed and why. Actual cost by work package is the best input to the next estimate, which brings you back to key three.
Pre-launch checklist
- Business case and scope approved, including what is out of scope.
- Acceptance criteria agreed for each major deliverable.
- Work breakdown structure complete to an estimable, assignable level.
- Schedule and budget approved as the baseline.
- Initial risk register in place and contingency derived from it.
- Sponsor named and approval thresholds written down.
- Responsibility matrix shared with the team.
- Change procedure known to the team and the client.
- Key resources booked and conflicts with other projects resolved.
- Project and work package codes set up in the system so costs post correctly from day one.
The last item is often missed. If spending starts before the project exists in the system, early costs land in general accounts and someone has to reallocate them by hand later, and they rarely all find their way back.
Mistakes that sink projects
- Starting work before scope and acceptance criteria are signed off.
- Estimates made by people who will not do the work.
- Contingency spread across packages and spent before it is needed.
- Verbal changes with no assessment or variation order.
- Progress measured by gut feel instead of planned, earned and actual.
- Project data scattered across spreadsheets and email, so no single reliable picture exists.
For a closer look at the obstacles, see our article on project management challenges.
Where an integrated system helps
Standalone scheduling tools show the plan but not what was spent. Accounting systems know what was spent but not what was done. An integrated ERP brings the two together: purchase orders, material issues and labor hours post to the project and work package as they happen, so budget versus actual is available at any time without manual consolidation.
Tranquil's project management solution tracks labor and parts costing by project, and its contracting pages cover cost and revenue estimation, daily activity management, material management, work in progress and project budgeting. If you run several projects at once, read our guide to multi-project management.
Frequently Asked Questions
What is the single most important success factor?
A clear scope with acceptance criteria from the start. Most later problems with schedule, cost and client relations trace back to ambiguity about what was agreed. Disciplined change control comes a close second.
How much contingency should a project carry?
There is no universal percentage. Build it from the risk register by adding up probability times impact for each risk, hold it at project level and release it by decision.
What is earned value in one paragraph?
It compares three figures: the budgeted value of work planned to date, the budgeted value of work actually done, and the actual cost. From them you get cost and schedule performance indices and a forecast of the final cost.
How should we handle client change requests?
Log them in writing, assess the effect on schedule and cost, and show the client that effect before doing the work. In contracting, do not execute a significant change without an approved variation order adjusting value or duration.
How can I tell early that a project is heading for trouble?
Watch three signals monthly: a cost index below 0.95, a schedule index below 0.9, and change requests piling up undecided. If two appear together for two months running, take the project to its sponsor for replanning, scope renegotiation or extra resources.
Does Tranquil track project costs?
Yes. Tranquil's project solution tracks labor and parts costs per project, with contracting modules that include project budgeting and work in progress. You can contact the Tranquil team to discuss how you manage your projects.