Sawmill Commerce
Most sawmill optimization conversations start at the saw. They're starting in the wrong place. By the time a log enters the line, the achievable yield is already mostly fixed — by the structure of the order book the planner was given to work with. Commerce is not the support function for production. Commerce sets the ceiling. Get it wrong and no amount of saw-line discipline recovers the gap.
Why commerce is the #1 lever
Commerce and planning together carry roughly 70% of the optimization potential before a single log is cut. This article zooms in on the first half of that: commerce.
The reason commerce sits at the top of the optimization stack is that it sets the menu the planner has to work with. Planning is a downstream function — it can only optimize within the constraints handed to it by the order book. A talented technologist with a clean, tolerance-aware, well-sized order book will find clever cant patterns and build full kiln charges. That same technologist with an order book of 47 unique specs, half of them sized for someone else's kiln package, will spend their week firefighting and deliver mediocre yield no matter how skilled they are.
The hidden cost of an over-promising sales process is that the costs are invisible at the quote stage. Sales teams are rewarded for closing orders, not for screening them. Every quote a sales rep sends out without a feasibility check is a potential margin landmine that detonates two weeks later, on the saw line, on a Monday morning, when the technologist realizes the spec doesn't fit the available equipment.
Three signs your commerce process is leaking margin
- More than 15 distinct specs cover 80% of monthly volume. This is the quickest indicator of spec proliferation. Mills with disciplined commerce typically run 8–12 specs covering 80% of volume; outliers run 20+.
- The sales team can't tell you the saw-spacing limit on the multi-rip without phoning the foreman. If a constraint that affects yield on every quote isn't on a one-pager next to the salesperson's monitor, it's not in the quoting process — it's in someone's head. That's a process gap.
- The technologist regularly "makes orders fit" on the morning of the shift. Last-minute pattern adjustments to accommodate a spec the planner didn't know about until Friday afternoon are not optimization; they are damage control. The cost shows up as yield loss, but the root cause is in commerce.
If any two of these are true at your mill, the rest of this article is a worked example of where the gap is hiding and how to close it.
Anatomy of a bad order — the 30×110 case
To make the costs of weak commerce concrete, walk through one anonymized case from a European softwood mill. The mill operates a single line with a multi-rip primary breakdown saw and an in-line splitter (resaw) for side-board recovery. Two technical constraints define what the line can do:
- Multi-rip minimum saw spacing: 32 mm. The saws on the multi-rip cannot be set closer to each other than 32 mm. Below that, the saw guides physically interfere. This is the minimum thickness the mill can produce on the primary breakdown.
- Splitter input thickness range: up to 27 mm. The splitter is a resaw that takes a thicker board on input and produces two thinner boards. Its mechanical input limit is 27 mm — boards thicker than that cannot pass through.
The sales team accepted an order: 30×110, 800 m³, at a price that looked attractive at the quote stage. The customer was a recurring client. The order entered the production schedule. Then the technologist tried to plan it.
Why the 30×110 had nowhere to go
Splitter route — closed. The 30×110 cannot be made by feeding a thicker side board into the splitter, because any input thicker than 27 mm exceeds the splitter's mechanical limit. 30 mm > 27 mm. The splitter route is physically unavailable.
Multi-rip / primary breakdown route — closed. To produce a 30 mm thick board on the multi-rip, two adjacent saws must be set 30 mm apart (plus kerf). But the minimum saw spacing on this line is 32 mm. The saws cannot get close enough to produce 30 mm thick boards. The multi-rip route is also physically unavailable.
Result. The 30×110 spec cannot be produced on this line by any clean route. The order ran through compromised cant patterns — patterns chosen to extract something close to the spec from logs that would have produced higher yield on a different cut. Yield on the 800 m³ batch dropped roughly 2.5 percentage points compared with comparable shipments. At an average value of €280/m³, that's ~€5,500 of margin lost on one order — for a customer who would almost certainly have accepted 32×110 or 30×100 if the salesperson had asked.
What makes the 30×110 case instructive is that none of the people involved did anything obviously wrong. The salesperson took a profitable-looking order from a known customer. The customer specified a dimension that's perfectly reasonable in the abstract. The technologist tried to make it work with the equipment they had. The yield loss was the cumulative result of a process gap: the salesperson didn't know about the 32 mm and 27 mm limits, and there was no checkpoint between the quote and the production schedule where someone could have caught it.
This is what the four practices below are designed to prevent.
The 4 commercial practices
The four practices below are not a maturity model — they're a dependency stack. Each builds on the one before it. Practice 1 sets the dimensional currency of the order book. Practice 2 reduces the number of distinct items in that currency. Practice 3 ensures the mix of items the planner can see at once is balanced. Practice 4 is the gate that prevents items from entering the book in the first place if the line can't run them.
Practice 1 — Size to the kiln package, then fill the chamber
The kiln is almost always the bottleneck of a sawmill. It runs on cycles, not on continuous flow, and the cycle length depends on board thickness, not on how full the chamber is. A 22 mm spruce charge takes the same number of hours to dry whether the chamber is loaded to 100% or to 70%. That means every percentage point of underutilization in the kiln is a percentage point of lost productive capacity for the entire mill.
Commerce can't fix kiln scheduling, but it can stop sabotaging it. There are two practices here.
1A — Round orders to multiples of your kiln package
Take a typical case: 22 mm thick × 100 mm wide boards, dried with 25 mm stickers between layers. The standard package on a typical European softwood line is roughly 4.0 m long × 1.2 m wide × 1.2 m high. The math works out as follows:
- Boards across the package width: 1200 mm ÷ 100 mm = 12 boards per layer
- Layers in the package height: board (22 mm) + sticker (25 mm) = 47 mm per layer; 1200 mm ÷ 47 mm ≈ 25 layers
- Boards per package: 12 × 25 = 300 boards
- Net timber volume per package: 300 × 0.022 × 0.100 × 4.0 ≈ 2.64 m³
Two practical consequences flow from this:
- Order volumes that are not multiples of 2.64 m³ leave half-empty packs. A half-empty pack is a half-empty kiln slot is lost capacity. Rounding orders to package multiples (or rounding up in the negotiation: "we'll deliver 2.64 × 12 = 31.7 m³, can we send the spare 1.7 m³ to your next order?") is the simplest fix in this entire article.
- The package geometry should be on a one-page cheat-sheet for sales. For each common spec — 22×100, 22×125, 32×100, 47×100, 50×150 — there should be a single number: "package volume is X m³, round orders to multiples of X." The salesperson doesn't need to derive this; they need to look it up.
1B — Backfill partial charges with adjacent-thickness specs
The kiln cycle is governed by the thickest board in the chamber. Mixing widths in a charge is fine — drying speed depends on thickness, not on width. Mixing thicknesses is not fine: thinner boards over-dry while the thicker boards finish, which means quality loss, energy waste, or both.
This creates a specific commercial opening. If the order book has a charge that's only 30% full of 22×100, the remaining 70% should be filled with orders of the same thickness — any width. 22×75, 22×125, 22×150 all work. 25×100 does not.
Practically, this means commerce should know which charges are currently underfilled and what thickness they need. The conversation between sales and the technologist is no longer "what should we pile in next?" but "we have a 22 mm charge sitting 70% empty for next week — what 22 mm orders can we recruit?"
The sales effect of this is concrete: it gives the salesperson a specific reason to call a customer who buys 22×125 occasionally and ask if they need anything in the next two weeks. That call closes 5–10 m³ at standard pricing and turns a half-empty kiln charge into a full one.
The two halves of Practice 1, in one rule
Headline orders set the dominant kiln charges. They should be sized as multiples of the package volume so they don't leave half-empty packs at the boundary.
Backfill orders finish the charges. They share the dominant order's thickness (within ±1 mm) and can be any width. Sales should hunt for them actively when a charge is undersubscribed.
Practice 2 — Consolidate adjacent specs within EN 1313-1 tolerance
Most sawmill order books carry far more distinct specs than the line actually needs to run. Two orders that look like two specs are often one spec separated by a couple of millimeters that fall well within the tolerance the customer's standard already allows. Running them as one spec is free yield.
The reference framework here is EN 1313-1 — Round and sawn timber: Permitted deviations and preferred sizes (Part 1: Softwood sawn timber). The standard defines two tolerance classes for sawn dimensions:
- Class 1 (precision): ±1 mm for dimensions ≤ 100 mm, and ±1.5% for dimensions > 100 mm
- Class 2 (standard): ±2 mm for dimensions ≤ 100 mm, and ±2% for dimensions > 100 mm
Most B2B softwood orders run to Class 2 in practice, even when the order doesn't say so explicitly. The customer's specification, if traced back to their construction or industrial standard, almost always references either EN 1313-1 directly or a national standard with equivalent tolerances. The 2 mm allowance in Class 2 is the working space commerce should be using.
The mechanics of consolidation
Walk through a concrete example. Two orders sit in the order book:
- Order A: 22×100, 200 m³, customer X
- Order B: 20×98, 150 m³, customer Y
The dimensional difference is Δt = 2 mm and Δw = 2 mm. Both deltas fall within EN 1313-1 Class 2 tolerance (±2 mm for dimensions ≤ 100 mm). Both orders can be consolidated and run as a single spec — produced at 22×100 — with both customers receiving timber that meets their stated standard.
The mill effects compound:
- One spec instead of two on the cant pattern. Fewer pattern changes per shift.
- One kiln package size instead of two. Cleaner kiln scheduling.
- Larger production batches. A consolidated 350 m³ order is much easier to slot into a single shift than two batches of 200 and 150 m³.
- Less work for the technologist. One specification to plan against, not two.
- Fewer setups, less downtime, more saw-line uptime. Each pattern change typically costs 15–30 minutes; eliminating one a shift recovers two hours of effective sawing per week.
How to have the conversation with the customer
Most customers don't know they have tolerance flexibility. They have a number on a spec sheet and they assume that's what they're buying. The conversation isn't "we want to sell you something different" — it's "what's the actual dimensional tolerance in your standard?" Once the customer confirms they operate to EN 1313-1 Class 2 (or the equivalent national standard), consolidation becomes a non-issue commercially.
A short order-confirmation language sample, suitable for inclusion in standard contract templates:
The consolidation lever does not require any customer renegotiation. It requires the sales team to not create distinctions that don't exist. A 22×100 order at 200 m³ and a 20×98 order at 150 m³ from two different customers should arrive in the order book as one entry: 22×100, 350 m³, two delivery addresses — not two specs.
Practice 3 — Mix thick and thin specs across the order book
The third practice is about the mix of the order book at any point in time. A balanced book — meaningful share of thick specs, meaningful share of thin specs — gives the planner room to design cant patterns that produce both at once. An unbalanced book — say, 80% thick and 20% thin — forces the planner into thick-board patterns where the thin orders fall out as awkward by-products.
The geometric reason is that side boards from a thick-board cant pattern are natural thin-board candidates. A cant pattern designed around 50×100 center boards leaves side boards on the outside of the cant — wood that has to come off the log somewhere, and that's most efficiently turned into thin boards (25×100, 22×100) sized to the side-board geometry. But this only works if there are thin orders in the book waiting for those side boards. If the book is all thick orders, the side boards become low-grade falldown — produced anyway, but with no buyer lined up at full price.
The corollary, when a thick spec dominates: actively recruit thin orders at widths that fall out of the dominant cant pattern naturally.
The splitter case — pair widths explicitly
Mills equipped with an in-line splitter (resaw) have a stronger version of this practice available. The splitter takes a thick side board on input and produces two thinner boards on output. With the splitter in the line, the planner can deliberately pair widths: a thick spec at one width, and a thin spec at a width that's geometrically half (minus kerf) of a side board produced from the thick cant.
For example, a 50×100 thick spec running through a cant pattern that produces 50×50 side boards on the outside — those side boards can be split into two 22×50 thin boards, if there's a 22×50 order in the book. Without the thin order, the side boards are just lumber. With the thin order, the same wood serves two customers and the yield on the cant pattern climbs noticeably.
This requires deliberate commercial action: the salesperson recruiting a thin order isn't filling pipeline — they're enabling the thick orders already in the book. Sales incentive structures rarely reward this kind of complementary recruitment, which is why it doesn't happen automatically.
The no-splitter case — thin specs need careful handling
Mills without a splitter face the inverse situation. Thin specs become harder to produce efficiently because every cant pattern must close cleanly across the entire cant, with no resaw step to recover side-board geometry. Thin orders that look profitable in isolation can be break-even or worse on the line because the patterns required to produce them leave more falldown than the patterns for thick orders.
Practical implication: in a no-splitter mill, thin specs should have a cap on their share of the order book — typically no more than 15–20% of monthly volume — and the cap should be enforced through pricing or order acceptance. Adding more thin orders past the cap erodes yield without proportional revenue gain.
Practice 4 — Accept orders against your real machine constraints
Practice 4 is the gate. The first three practices optimize the order book that gets through. Practice 4 prevents orders from getting through if the line can't run them. It is what failed in the 30×110 case at the start of this article.
The principle is that a salesperson should never quote a spec without checking it against a written list of the line's hard physical limits. The list exists. The salesperson reads it. If the spec falls inside the limits, the quote goes out. If the spec falls outside, the salesperson either renegotiates with the customer or routes the order to the technologist for a feasibility decision before quoting.
The five constraints every commercial process should screen against:
- Multi-rip minimum saw spacing. The minimum gap between adjacent saw blades on the primary breakdown saw. Sets the minimum thickness producible on the multi-rip. Typical range: 25–32 mm depending on saw guide design.
- Splitter (resaw) thickness range. The minimum and maximum board thickness the splitter accepts as input. Sets which orders can use the splitter route at all. Typical range: 18–27 mm input, varies widely by machine.
- Stacker downstream of the splitter. The stacker (the package-forming line that receives boards after the splitter and assembles them into kiln packages) typically caps the share of splitter-bound patterns in any given week — its throughput, not the splitter's throughput, is the real downstream constraint. Practical cap: 30–40% of weekly volume on splitter-bound orders.
- Edger working range. Minimum and maximum board width that the edger can process in a single pass. Specs outside this range either can't be produced or require additional handling.
- Trimmer length grid. The discrete lengths the trimmer is set up to produce cleanly. Order specs that don't fall on this grid lose 5–10 cm to trim per board, which compounds across a 200 m³ order into measurable yield loss.
Each of these constraints has a number for your specific mill. The number doesn't need to be derived; it needs to be looked up. The cheat-sheet in the next section is the form that lookup should take.
The 30×110 case, retold against the cheat-sheet
If the salesperson on the 30×110 order had checked the cheat-sheet at the quote stage:
Constraint 1 — Multi-rip min spacing 32 mm. Spec calls for 30 mm. Flag: red. Cannot be produced on the multi-rip.
Constraint 2 — Splitter input ≤ 27 mm. Spec calls for 30 mm input. Flag: red. Cannot be produced via the splitter.
Verdict at the quote stage: "Both production routes are blocked. Renegotiate to 32×110 or 30×100, or refer to the technologist before quoting." The €5,500 of margin loss never happens, because the order never enters the production schedule at 30×110 in the first place.
The constraint cheat-sheet
The cheat-sheet is a single page, A4, in landscape, laminated, taped next to the salesperson's monitor or kept in a shared file the sales team opens before every quote. It is not a training document — it is a reference document. Salespeople do not need to remember the numbers; they need to look them up.
Below is a template version. Each mill fills in its own values. The technologist owns the document and updates it whenever a machine setting changes.
| Constraint | Your mill's value | When to flag |
|---|---|---|
| Multi-rip minimum saw spacing | e.g. 32 mm | Specs requiring board thickness below this value cannot be produced on the primary breakdown. |
| Splitter input thickness range | e.g. 18–27 mm | Side-board input outside this range cannot pass through the splitter. Specs that need to be produced via the splitter route must fall inside it. |
| Stacker share of weekly volume (splitter-bound) | e.g. ≤ 35% | Cap on the share of weekly volume that depends on the splitter-and-stacker route. Above this, the stacker becomes the bottleneck. |
| Edger minimum board width | e.g. 75 mm | Specs below this width require additional handling or cannot be produced cleanly. |
| Trimmer length grid | e.g. 3.0 / 3.6 / 4.2 / 4.8 / 5.4 m | Spec lengths off the grid lose 5–10 cm to trim. Negotiate to grid lengths in the quote. |
| Common kiln package volumes | e.g. 22×100 = 2.64 m³ / pack; 22×125 = 3.30 m³ / pack; 32×100 = 3.84 m³ / pack | Order volumes should be multiples of the package volume for the spec. |
The cheat-sheet is the simplest, lowest-cost commerce intervention available. It does not require new software, training programs, or personnel changes. It requires the technologist to write down five numbers and the sales team to read them.
From practice to process — operationalizing commerce
Knowing the four practices doesn't change anything. Wiring them into the workflow does. The translation from practice to process happens through four artifacts and one cultural intervention.
The four process artifacts
- The constraint cheat-sheet (above). Owned by the technologist. Read by sales before every quote. Updated whenever a machine setting changes — typically once or twice a year.
- A feasibility check step in the quote workflow. Before any quote above a threshold (e.g., > 100 m³, or any spec not previously run by the mill) is sent to the customer, it goes through a 5-minute screen against the cheat-sheet. The screen produces a green/amber/red flag. Green: quote proceeds. Amber: quote proceeds with a note to the technologist. Red: quote does not proceed until the technologist has confirmed feasibility.
- A weekly sales/operations sync. Thirty minutes a week. Sales walks the technologist through the live pipeline and the next two weeks of expected orders. The technologist flags anything that won't run cleanly and identifies kiln charges that need backfilling. The output is a short list of "things to recruit": specifically which thin orders are needed against the dominant thick orders, which 22 mm orders are needed to fill an undersubscribed kiln charge, and which off-grid lengths need to be renegotiated.
- A "specs covering 80% of volume" KPI. Tracked monthly. Trended down. The KPI lives where the sales team can see it and is owned by the head of commerce, not by the technologist. The target is typically 8–12 specs covering 80% of monthly volume; mills running above 15 have visible spec proliferation.
The cultural intervention
Process artifacts work or fail based on whether the people involved actually believe they matter. The most reliable way to make the sales team believe is to put them on the production floor.
How to actually fix the sales/production gap
A. Send the sales team to work on the production floor for two or three days.
Not a tour, not a presentation in a conference room — actual shifts at the saw, the sorter, the kiln, the stacker, the trimmer. Each salesperson observes a full production day, ideally including a pattern changeover, a partial kiln charge, and at least one moment where the technologist has to make a spec fit something it doesn't quite fit. The first time a salesperson watches a 30-minute pattern change triggered by a 4 mm difference they negotiated, the way they quote changes. This is a one-time investment that resets the quoting culture for years.
It also builds relationships. The salesperson and the technologist now know each other. The Friday-afternoon phone call to ask about a marginal spec stops feeling like an interruption and starts feeling like a normal part of the workflow.
B. Sales cannot accept orders without production sign-off.
Make this a hard process rule, not a guideline. Every order above a defined threshold — for example, > 100 m³ on any single spec, or any new spec not previously run by the mill, or any spec flagged amber/red on the constraint cheat-sheet — requires explicit sign-off from the technologist before the quote is sent to the customer.
The bottleneck this creates is the point. It forces the conversation that would otherwise happen on the saw line at 6am on Monday to happen at the quote stage on Tuesday afternoon, when there is still time to renegotiate or decline. The friction is mild compared with the alternative.
Together, the four process artifacts and the cultural intervention typically take two to four weeks to install and pay back inside one quarter — measurably, in yield numbers and changeover hours, not in soft metrics.
For the planning side of this — turning the consolidated, constraint-checked order book into optimal cant patterns — see sawmill production planning. Commerce builds the menu; planning chooses the dish.
What to do this week — checklist
Five diagnostic actions that surface the gap and point to the next step. Pick the one that exposes the largest gap at your mill and act on it this week.
Sawmill commerce — week 1 checklist
- Pull last month's order book. Count the distinct specs that cover 80% of volume. Above 15 — there is a spec proliferation problem worth tackling immediately. Above 20 — Practice 2 (consolidation) will pay back in one quarter.
- Open your last 10 quotes. How many would have been flagged by Practice 4 (machine-constraint check) if a cheat-sheet had existed? Even one is a process gap that cost or will cost yield.
- Walk to the kiln. Pick a charge in progress. Measure the gap between the top of the pack stack and the false ceiling, and from the side of the stack to the wall. Above 5 cm on either dimension means Practice 1A is being violated upstream — package volumes don't match the chamber.
- Ask your top 5 customers (by volume) which tolerance class their spec actually requires. Most won't know off the top of their head. That's your consolidation opening. Reference EN 1313-1 Class 2 in the conversation.
- Check whether sales has a written constraint cheat-sheet within reach of their desk. If not — the technologist has 30 minutes of work to write one. That single artifact is the highest-leverage commerce intervention available.
Where to go next
Two ways to act on this
If your spec count is above 15 covering 80% of volume, or you can name an order in the last quarter that ran into a constraint at the saw line: the gap is process, not capability. The fastest move is a structured 30-minute commerce audit — we walk through your last quarter's order book, identify the consolidation opportunities and the constraint flags that should have been caught at quote stage, and hand you a draft cheat-sheet for your line.
Sources & further reading
The European tolerance and packaging standards referenced in this article are the operational backbone of softwood commerce in the EU. Anonymized client data is based on real project work in European softwood sawmilling.
Standards (most relevant to this article)
- EN 1313-1 — Round and sawn timber: Permitted deviations and preferred sizes (Part 1: Softwood sawn timber). Defines Class 1 and Class 2 dimensional tolerances referenced throughout Practice 2 of this article. cen.eu / iso.org
- EN 336 — Structural timber: Sizes, permitted deviations. Defines target sizes and tolerances for structural-grade softwood and hardwood, used widely in construction-grade orders. cen.eu
- EN 1611-1 — Sawn timber: Appearance grading of softwoods (Part 1: European spruces, firs, pines, and Douglas fir). Defines visual grading classes and defect allowances — relevant to Practice 2 conversations about grade tolerance with customers. cen.eu
- EN 14081-1 — Strength-graded structural timber with rectangular cross-section: General requirements. Reference standard for strength-graded products that downstream order specifications inherit from. cen.eu
- EN 14298 — Sawn timber: Assessment of drying quality. Reference standard for kiln-dried product quality, relevant to backfill discussions in Practice 1B. cen.eu
Industry data & market context
- Eurostat — Sawnwood production statistics (PRODCOM). Quarterly production volumes and turnover for EU sawmills. ec.europa.eu/eurostat
- FAO — ForeSTAT / FAOSTAT Forestry Production and Trade. Global softwood and hardwood production, trade, and consumption time series. fao.org/forestry/statistics
- EOS — European Organisation of the Sawmill Industry, Annual Report. Production, capacity utilization, and outlook for European softwood sawmills. eos-oes.eu
Operational practice & technology
- FPInnovations — Lumber recovery and sawmill optimization research. Peer-reviewed research on cant pattern optimization, log breakdown strategies, and yield modeling. fpinnovations.ca
- USDA Forest Service — Forest Products Laboratory technical reports. Reference data on kiln drying, lumber grading, and recovery economics. fpl.fs.usda.gov
- Skogforsk — Forest research on log sorting, scaling, and bucking optimization. Practical research from Nordic sawmilling. skogforsk.se
- Wood Industry / International Forest Industries / Timber Processing. Trade publications covering sawmill operations, optimization technology, and case studies. woodindustry.ca · internationalforestindustries.com · timberprocessing.com
The 30×110 case and the 28-to-11 spec consolidation result referenced in this article are based on real project work with European softwood sawmills. Specific client identification has been omitted to preserve commercial confidentiality. Numerical values (yield deltas, margin impact, machine constraint values) are representative of typical European softwood operations and may vary by mill configuration.