Why Do Renovation Projects Always Overrun?

Because the plan assumes the building is as surveyed, the trades are available when needed, and the client decides once. In practice all three assumptions fail, and each failure compounds the next rather than replacing it.

Overruns are not usually caused by slow work. They are caused by work that starts before the information exists, and by decisions that arrive after the material was ordered.

What is the most common cause of delay?

Discovered conditions after demolition: asbestos, failed structure, illegal wiring, drainage that does not connect. Each stops work and starts a variation.

The second is late decisions. A tile chosen in week three rather than week one moves the tiler and everything behind them.

Why do discoveries always happen?

Because a survey samples the building rather than inspecting it entirely. Nobody removes a floor to check a joist unless there is a reason.

The remedy is not a more thorough survey, which has diminishing returns. It is a contingency in money and time that assumes at least one discovery.

How much do decisions delay a project?

A late decision on a finish costs a week; a late decision on a service layout costs a fortnight. Services have to be decided before first fix, not before second.

Decide in the order the work happens, not in the order of interest. The kitchen colour can wait; the socket positions cannot.

What role does the client play in overruns?

A large one. Changes of mind during the build are the single most controllable source of delay, and they are almost always presented as small.

Every change reopens a stage, and reopening a stage costs the remobilisation of a trade, not just the change itself.

Why do trades arrive at the wrong time?

Because they are booked by availability rather than by sequence, and because the previous trade finished late. The two compound.

Booking the second fix before the worktop template is confirmed is how a plumber ends up standing in an empty room.

How does weather affect an internal renovation?

Through the envelope. A roof opening, a new extension or a window replacement is weather-dependent, and a wet week can stop plaster drying for a further week.

Programme envelope work for the drier months where the climate allows. It is the cheapest scheduling decision available.

What is the real cost of a one-week delay?

The trade’s standing time, the extended scaffold hire, the extended site insurance, and the alternative accommodation. On a typical project that is 500 to 2,000 a week.

Delay costs are rarely in the original budget, which is why they feel disproportionate when they arrive.

Can a project be compressed to recover time?

Rarely without cost. Crashing a schedule means overlapping trades that should be sequential, which produces rework and callbacks.

The one legitimate compression is pre-ordering long-lead items early. Everything else trades money for time at a poor rate.

What are long-lead items?

Kitchens, windows, bespoke joinery, stone worktops, steel beams and anything imported. Lead times of six to sixteen weeks are normal and they are the real critical path.

Order them before demolition, not after. A project waiting for a kitchen is a project paying standing time.

How should the programme be built?

Backwards from the completion date, with the long-lead orders placed first, the inspections booked second, and the trades sequenced third. Then add the contingency at each end.

A programme built forwards from today ignores the constraint that actually governs it.

What is the most reliable predictor of overrun?

The number of decisions still open at the start of the build. Above a handful, the project will slip, and the slip is proportional to how many services are affected.

Close the services decisions first. They are the ones that cannot be changed later without opening a wall.

How do you handle a client change mid-project?

Price it, programme it, and get it signed before the work starts. A change accepted verbally is a change that will be argued about at the final invoice.

Most contractors will accommodate a change if it is priced properly. The problem is the change that arrives as a preference rather than an instruction.

Does a larger contractor reduce overruns?

Not necessarily. A larger firm has more resources and less flexibility, and the site team changes more often. A small firm with a good programme often beats them.

The predictor is the quality of the programme and the references, not the size of the company.

What about the last five percent?

It takes as long as the first fifty. Snagging, commissioning, certificates and the final decoration are slow because they depend on other people’s diaries.

Programme four weeks for the tail of a project and expect to use them. Projects that promise a two-week tail rarely finish in four.

How should the household plan around it?

Assume the kitchen is unusable for two weeks longer than promised and the bathroom for one. Temporary arrangements that assume the best case fail in the worst week.

Plan the disruption for the realistic duration, not the optimistic one. It costs nothing and prevents most of the stress.

Can a penalty clause help?

For a large contract, yes, if it is realistic and reciprocal. An unenforceable penalty is worse than none, because it encourages the contractor to walk away instead of finishing.

Liquidated damages need to reflect actual loss. A round number invented at signing is usually unenforceable.

What is the one habit that reduces overrun most?

A weekly site meeting with the programme on the table and the next two weeks’ decisions confirmed. It is boring and it works.

The meeting forces the decisions forward and exposes the slippage while it is still recoverable.

How does the number of trades affect the risk of overrun?

Directly. Five trades mean four handovers, and every handover is a day that can be lost. Ten trades mean nine opportunities for a gap.

That is why a package approach, where one firm covers several trades, often runs closer to programme than the cheapest split.

Can a written programme prevent overrun?

It prevents the avoidable part. A programme with dates, inspections and long-lead orders removes the uncertainty the client controls, which is most of it.

It cannot prevent a discovery behind a wall. That is what the time contingency is for.

What should be accepted as normal?

One to two weeks of slippage on a two-month project, and three to four on a four-month project. That is the baseline, not a failure.

Planning for the baseline is what separates a controlled project from an optimistic one. The guide on contingency and the one on payment stages both assume it.

Overruns are a systems problem rather than a discipline problem, which makes them partly preventable. The guides on contingency and on long-lead ordering cover the two levers that work.