Why projects fail before they start

Many troubled projects were weak before delivery began. The problems were built in when the money was committed: optimistic benefits, missing costs, untested assumptions and supplier contracts that rewarded the lowest price over deliverability. The good news is that these are among the cheapest problems to fix — provided they are found early.

What a credible business case contains

• A clear problem statement: what is not working today and what it costs.
• Options, including ‘do minimum’, so the preferred choice is genuinely compared.
• Measurable benefits, with an owner and a date for each one.
• Whole-life costs: delivery, running, support, training and the time of internal staff.
• Key assumptions and the risks that would undermine the case if they proved wrong.
• A realistic plan and the capacity needed to deliver it.

Buy outcomes, not just the lowest price

Procurement shapes a project for years. Specify the outcomes you need, not only the tasks. Weight quality and delivery approach alongside price. Ask suppliers how they would handle the main risks, and check the experience of the people who will actually work on your contract. Build in clear acceptance criteria, regular performance reviews and a fair way to manage change.

Keep accountability in-house

Suppliers can be accountable for what they deliver, but the organisation remains accountable for the result. Name an internal owner for the outcome and the benefits, and make sure they have the time and authority to hold the work together.

A typical scenario

In an illustrative case, a business case for a new system shows a two-year payback. An independent review finds it excludes training, data migration and the time needed from operational staff. With those costs included, payback moves to three years — still worthwhile, but the plan now funds the work that would otherwise have been squeezed, and the risk of a mid-project funding gap is removed.

Before you commit: checklist

• Would the benefits still justify the investment if delivery took 25% longer?
• Have all running and people costs been included?
• Has each key assumption been tested or given a date to test it?
• Do supplier contracts define what ‘done’ means?
• Is there a named internal owner for the outcome?

A short independent challenge before commitment costs a fraction of a struggling project, and it protects everything that follows.