Case study

Oak Bank care home

A live care home build for Meridian Construction, replanned around zones and promises while it was running.

Meridian are a regional contractor who build well and win on reputation. The problem was never the building. It was that the programme and the site had stopped being the same thing, and nobody could say so early enough to act on it.

Care home · live build Takt and the Last Planner System Named with Meridian’s permission

Book a 30-minute call

What was happening

The job had a programme. It was a good programme, built properly at tender. By the time the first fix started it described a building that no longer existed: dates had moved, trades had been resequenced on the hoof, and the version on the wall was two revisions behind the version in the planner’s file.

What followed is the pattern on most jobs in trouble. Trades turned up to areas that were not ready and went away again. The site manager held the real plan in his head and on his phone. Problems reached the directors as recoveries rather than as decisions — by which point the only options left were the expensive ones.

None of that was a people problem. Every one of those behaviours was a rational response to a plan nobody could trust.

What we changed

Two things, in this order, because the second does not hold without the first.

Promises first

The Last Planner System went in before anything was redrawn: a pull plan built backwards from the milestone with the trades in the room, a six-week look-ahead, and a constraint log with a name and a date against every entry.

The change that mattered was to the weekly meeting. It stopped being a progress report and became a set of promises, made by the people who would do the work. Then we counted how many were kept, every week, in front of everyone.
Then the pace

We split the building into zones holding roughly equal work — drawn from the work in them, not by cutting the floor plate into equal squares — put the trades in sequence as wagons, and agreed one beat that every zone handover would be measured against.

From that point the question at the weekly meeting changed. It was no longer “are we on programme?”, which nobody could answer honestly. It was “did zone three hand over clean on Friday?”, which everyone could.
And a look at the business behind the job

Some of what we found on site was not about the site at all: how work was handed from pre-construction to delivery, who could actually approve a variation, and how long information took to reach the people holding the drawings.

That strand of the work is what became Blueprint. Oak Bank is where it became obvious that fixing one job well is worth a great deal, and that the same job comes back around unless the business behind it changes too.

What we measured

Three numbers, tracked weekly from the first pull plan. They move before anything shows on the programme, which is the point of them.

Promises kept

Every week, each trade says what it will finish. The following week we count what actually finished, and divide one by the other. That single number is the most honest thing on a construction site, because nobody can argue with their own commitment from seven days ago.

~50% → ~80%

Promises kept on Oak Bank, over roughly eight weeks

The first count is always the uncomfortable one, and Oak Bank was no exception: about half of what had been promised the week before did not happen, on a job everyone believed was going reasonably well. That gap is not laziness. It is a plan nobody could have kept, and until somebody counts it nobody knows it is there.

Eight weeks later it was running at about eighty. Nothing was added to get there: same trades, same drawings, same dates on the contract. What changed is that the people doing the work started saying what they would finish, and somebody started clearing what would have stopped them while there was still a week to do it in.

Those two numbers are the argument for this whole way of working. At fifty, every week quietly contains a second week of waiting and going back over things. At eighty, the plan is something the site can actually build to.

Figures from Meridian, rounded. They are happy to talk to anyone seriously considering this — ask on the call.

Clean zone handovers

Whether a zone went to the next trade complete, on the day it was promised, with nothing left to come back for. A zone handed over nearly finished is not finished: it is a return visit, a second mobilisation and an argument, booked in advance.

This is the number that tells you whether the beat is real or just drawn.

Days to clear a roadblock

From a constraint being raised to it being gone. Not how many were raised — how long each one sat there.

It is the only one of the three that measures the office rather than the site, and it is usually the one that stings. A site cannot plan around a drawing answer that takes three weeks, however well it plans.

What Meridian kept

A zone map and a wagon list for the remaining phases
The weekly planning meeting, run by their own team
A constraint log with an owner against every entry
Three numbers the directors can read in a minute

What we would do differently

Start earlier. We arrived with the job already running and the programme already adrift, so the first fortnight went on rebuilding trust in a plan rather than on building the plan itself.

Going in at the start of a phase rather than the middle of one is worth more than any amount of cleverness afterwards.

References

Meridian will take the call

Written and verbal references are available on request, and Meridian are happy to speak to anyone seriously considering this. Ask on the call and we will put you in touch.

Not sure if this is the right one? Let’s talk.

You tell me what keeps going wrong. I’ll ask how your jobs run from win to handover. Then I’ll tell you straight which of these is right for you, if any.

30 minutes, by video call
With me, not a sales team
No slides, no pitch

Pick a time that suits you

The calendar couldn’t load here.