The method · analysis, action, results
Seven steps between the figures you already have and a number that moved. Each one says what goes in, what comes out and how long it takes. Four of the seven you can run without us.
Before anything
A method that works everywhere works nowhere. The gate comes first, it is short, and a no at the gate costs you nothing. We expect to turn away about a third of the plants that ask.
Can you change something on the floor on Monday without asking anyone? If the decision sits with a parent company, a board or a partner who is not in the room, formation will stall at the first package redesign and the money is wasted.
Has the curve actually gone flat? Not a bad quarter. Growth or margin that stopped responding to the things that used to move it — the automation, the consolidation, the lean programme.
Is there work you could take if labour and capital stopped being the constraint? Existing customers who could give you more, or adjacent work you keep declining. Capability released into no demand is just a quieter plant.
There is a fourth, read rather than asked: the ground gate from the Bloom Readiness Assessment. Four indicators out of twenty. Below eight, the answer is not this year, and we will tell you what has to be true first.
Revenue from where it is to where it could go if labour and capital were not the constraint. Margin likewise. And the period you will give it.
This is the one thing we will not supply. A target we set is a consultant’s number and you will spend the year arguing with it. A target you set is the thing everything downstream gets measured against.
Alongside it, three figures you already have: what turnover costs you, what overtime costs you, what rework costs you. Those three put a floor under the conversation before anyone has an opinion.
Twenty indicators across the five levels at which value is taken out of a business before it reaches the people who made it: the instruction card, the dashboard, the pay stub, the package and the valuation. Zero to five each, twenty per level, a hundred in total.
Above eighty, built to last. Sixty to seventy-nine, drifting. Forty to fifty-nine, extracting. Below forty, built to extract. The total matters less than the shape: any single level under eight out of twenty is a break, and a break upstream makes everything downstream cosmetic.
You score it yourself. Nothing is collected, nobody audits it, and an inflated score only costs you the year.
The Siphon says what is draining. This says what the ground will hold. The ground gate first, then four elements — the mud a person can fail in, the water that forms them, the sun that tests them, and the seed corn that is set aside — twenty each, eighty in all.
Sixty-four and above, plant now. Forty-eight to sixty-three, plant in one place and watch it. Thirty-two to forty-seven, fix the ground first. Below that, nothing you plant will hold.
As with the Siphon, any element under eight starves the other three. Planting order is fixed: mud, then water, then sun. Formation without a place to fail safely produces compliant people, not capable ones.
The two models are not alternatives. Each element of the Bloom answers exactly one level of the Siphon, because the Siphon was built from the top down and the Bloom is planted from the bottom up.
The mud answers the instruction card. The water answers the dashboard. The seed corn answers the pay stub. The sun answers the package. The ground answers the valuation. Lay the two scores side by side and the flags stop being two lists of complaints and become one list of paired problems, each with the intervention that addresses it already attached.
A Siphon break with a healthy Bloom element beneath it is a policy problem and can be fixed in a quarter. A Siphon break sitting on a starved element is a formation problem and will take a year. Knowing which is which is most of the value of the day.
Select the frontline leader. Redesign the package so that what you say you want is what the package pays for. Redirect the cash into the capability account. Change the dashboard so it reports what the floor can act on. Return the instruction card to the people doing the work.
Owners almost always want to start at the end, because returning the card is visible and cheap and feels like trust. Done first, it fails: people hand the card straight back, because the dashboard still measures the old thing and the package still pays for it. You get one attempt at this with a given crew, and a failed attempt costs you three years of credibility.
The sequence is also where the afternoon on the floor earns its keep. The order is general; which of the five is already half-done in your plant is specific, and you cannot read it from a spreadsheet.
Hiring, vocational, accreditation, contribution. The first three are investments in a person and are amortised across three years at roughly a third a year; the fourth is what they return. Formation at entry runs around a quarter of paid hours and settles near a fifth averaged over three years — which is, not coincidentally, what the German dual system has cost for a century.
The worksheet ends in the restructuring test: if you would lay this person off in a downturn, you have not been investing in them, you have been renting them, and the account you are about to open is a bonus scheme with a better name.
The sequence, costed, laid against the A-to-B and the C-to-D from step one, at your own margin rather than a benchmark. Not a promise — a projection you can argue with, which is the only kind worth having.
Alongside it, the Capability Capital Score: eight drivers a buyer is already pricing whether or not anyone has named them. Bench depth, recorded capability, the formation engine, formed leaders, retention of the formed, capacity you already own, demand you could absorb, and independence from you.
The arithmetic is the part owners tend to sit up for. Capability released has a minor effect on EBITDA and a major effect on the multiple, because what a buyer is buying is an assembled workforce that is not separately recognisable on any balance sheet and lands in goodwill after the sale. One turn of the multiple is usually worth more than the whole year of operating gain.
One number was chosen at the end of the day, by you, and it was chosen because it would move within a quarter if the first intervention held. At ninety days we read it the same way it was read the first time, and we write down whether it moved.
If it did not, that is a finding and not a failure, and the usual cause is sequence rather than effort. At twelve months the two instruments are run again by the same hands. Compounding, if it is happening, looks like a score that rises in a year when nobody from outside was in the building.
What this is not
The gate, your figures and both reads are yours to run, free, in an evening. We have published the instruments rather than keeping them, because an owner who has read his own plant arrives at the sequencing conversation as a peer instead of a client.