At a 7% margin one lost ruble costs fifteen rubles of revenue. To recover ₽100k of losses you have to sell ₽1.4m. That is why stopping the leaks is almost always faster and cheaper than selling more.
Turnover holds or grows, yet there is no cash in the account. Nobody can say exactly where the profit goes.
The shop asks for new equipment, while you suspect the existing machines idle half the shift — and have no way to check.
You are short of people and there is nobody to hire. Yet no one has measured what anyone actually does during a shift.
The warehouse is packed with WIP and materials. The working capital you badly need is frozen in it.
Cost by product and by line is estimated “roughly”. Where you earn and where you lose is unknown.
You have tried improvement before. Six months after the consultants left everything was back where it started.
Each direction can be taken separately. Together they form a loop that holds the result after we leave.
Classic lean, but locked in digitally. I map the value stream, build the VSM and a waste map, and measure operation times. The output is a number in rubles and an honest answer on whether this process is worth touching at all. Then we rebuild one flow end to end and run PDCA cycles.
Where a human relay of information used to sit, an agent takes over. Models are deployed on premise, inside your own perimeter — no cloud, no data leaving the building. We write the software ourselves, so we own the result end to end.
I do the economics myself: build the P&L, break down unit economics, model the business case for the direction. Without that, any optimisation is guesswork. Separately I work on releasing working capital: at current interest rates, money pulled out of the warehouse is cheaper than any loan.
This is not about “cutting headcount” but about understanding who actually does what and how many people this volume really needs. With the current talent shortage that matters more than layoffs: it usually turns out that twenty-five people do the work of forty, while the rest compensate for someone else's chaos.
The most underrated part. As long as improvements are handed down from above, they last exactly as long as someone watches them. When an employee submits a proposal from the shift, sees what happens to it and gets paid for it, the mechanism runs without a supervisor.
Five days on site. A waste map priced in rubles and a set of priorities. If it is not worth doing, we say so plainly.
5 daysOne isolated area. If it does not work, rollback takes days and the rest of the plant never notices.
3–6 monthsMonitoring, dashboard, standards in the phone — the things that stop it sliding back.
in parallelInternal trainers, a development unit, procedures. After a year you can choose not to renew.
onwardA private club for business owners: live offline case reviews in central Moscow, a table assembled around your request, business games and access to a team of practitioners. Everything discussed stays inside the circle.