Decide the allocation before you chase the savings

Owning the mills is not the same as knowing where the wood should go

An 8 to 10 week sprint, February 2026. A €9B global forestry company. One senior analyst and two analysts, 20+ stakeholder interviews. Capability: Merger and Acquisition.

The company had recently integrated several sawmills with an existing downstream mill. On paper the structure was finished: material could move from the group's own sawmills into the group's own board operations, and the value of each transfer would stay inside the business instead of leaving it.

The decisions, though, were still being taken site by site, with sourcing and logistics set against local priorities. There was limited visibility into how material moved across the combined operation, and no single view of what any one routing was worth against the alternatives.

Three things followed from that gap, rather than from anything being badly run.

1. The fibre balance was inefficient, meaning the mix of grades reaching each destination was not the mix that paid best.

2. Procurement cost more than the structure required.

3. Value went missing in the choice between using material inside the group and selling it outside.

The integrated structure existed. The unified view across sites and sourcing that would let it pay did not.

They had run sawmills for decades and integrated them once

This is the point where a reader is owed the obvious question. Why did they not just do this themselves?

Concede the first half, because it is true. Everything this analysis needed was already inside the building: the volumes, the costs, the mill capacities, the contracts, the logistics lanes. No competitor held any of it, and no customer had to be interviewed to find it. On the inputs the company was fully equipped, and its operators understand those mills better than any outsider will.

The gap was repetition rather than knowledge. Running a sawmill is a skill built by doing it several thousand times, and this company has that repetition behind it. Working out which grade of material should go to which destination across a newly combined group, and what each choice is worth, is a task that arrives once, at integration. There is no second one to get right, and no previous version of their own to improve on, because the structure being allocated across did not exist until recently.

So the trade here is narrow and worth stating plainly. The company was short of practice at a job that comes up only when the structure changes, not short of understanding of its own business.

The group can now price a routing decision before it takes it

What that enables. Before the sprint an allocation decision was made locally and defended locally, and nobody could say what the group gave up by taking it one way rather than another. After the handover that decision carries a price before it is taken, and the price can be recomputed when timber costs, board demand or external prices move. That is a standing capability rather than a one off answer, because the inputs never stop moving.

The recommendation was the setup itself

The engagement produced no separate list of recommendations, and none has been invented here. That absence is the recommendation.

The distinction matters more than it sounds. A list of recommendations is a judgement made once, on one week's prices, by people who then leave the building. A setup with a model under it is a rule that can be re-tested every time the inputs move, by the people who stay. Only one of those two survives the first change in timber cost.

This matters to anyone with an integration of their own in progress. A recommendation tells you what to do once. A setup lets you ask again in six months.

€10m identified is not €10m banked

The €10m is an uplift identified during the sprint. That word is doing real work: the sprint found the opportunity and quantified it, and the company has not yet taken it. Taking it means running sourcing and allocation to the defined setup across a full cycle, which is the client's work rather than ours.

The 4 synergies were evaluated, which likewise describes work done rather than a result collected. Neither figure has been converted into a percentage or a run rate, because the engagement measured neither.

Source. SprintlyWorks case record, Optimizing Value in Integrated Sawmill Setup, a €9B global forestry company, February 2026, an 8 to 10 week sprint with 20+ stakeholder interviews. Both figures are the record's own, with its qualifications reproduced unchanged: 4 cost and revenue synergies evaluated, €10m EBIT uplift identified. The client is described by revenue scale and is not named, and nothing appears here about its sites, its counterparties, its commercial positions or any option it was weighing.

One question, 8 to 10 weeks, a model you keep

We take the business critical question a team cannot get to and answer it in 8 to 10 weeks, with one senior analyst and two analysts. On a question like this one the deliverable is a model the client owns and can run again, not a document.

The method is the part worth buying once. The model is the part worth keeping. The question to ask before scoping anything is whether your own team, given six uninterrupted months, would have the answer. If the reply is yes, you do not need us.

If you are inside an integration and the allocation rule has not been written yet, that is the conversation. Tell us the question: rahul.abhisek@sprintlyworks.com. If a sprint is the wrong tool, we will say so.