# Housekeeping Closet Organization Guide **Version:** 2.1 **Date:** 2026-08-21 **Supersedes:** v2.0 (2026-08-20). v1.0 (2026-06-15) stated the method with no building attached and remains available. **Type:** Method, with the property it was built on and the conditions it transfers under --- ## What changed in 2.1, and why it is printed here **Version 2.0 stated a fact about a second property that was not true.** It described a named property as scheduled to open, with "no status reports, no legacy worksheet, and no accumulated habit". That property's operating register showed it trading, and had been for some time. The sentence was wrong when it was published. The property is not named here and its trading figures are not reproduced. The correction is the point; the other building's numbers are not this document's to publish. Two things follow, and both are corrections rather than edits. **First, the claim is withdrawn**, not repaired. Section 7 no longer describes any second property's condition, because this document is not in a position to describe it — the transfer conditions belong to whoever holds that building's register, and stating them from outside is how the original error happened. **Second, section 7 now names no property at all.** A method document can name the building it was *built on*, because that building's own survey is the evidence. It should not name the building it is *carried to*, because that is a claim about someone else's operation made without their register in hand. The guide's own section 4 argues that a document showing only the corrected number is indistinguishable from one that guessed right. That applies to this document. The wrong sentence is quoted above rather than removed quietly. --- ## Why this version names a building Version 1.0 described how to organize a floor supply closet and named no property. It was correct and it was weightless. Any guide can assert that shelving should be standardized; what makes the claim worth anything is a building where it was actually done, a survey that shows the constraints, and a number that was wrong before the work and right after it. This version is the same method with the evidence attached. The property is **Westmark Fairbanks Hotel & Conference Center**, 400 rooms, tall tower floors 1–8. The drawings are its real closets, surveyed by hand on 15 June 2026. The error described in section 4 is a real error that was live on the floor. The last section carries the method forward to a property that has not opened yet. That is the actual test of a method: not whether it explains what already happened, but whether it can be run before the building has any history to read. --- ## 1. The zone principle Divide every closet into four fixed zones, always in the same order, on every floor: 1. **Linen** — sheets, pillowcases, duvet covers, mattress protectors 2. **Terry** — bath towels, hand towels, washcloths, bath mats 3. **Guest amenities** — soap, shampoo, conditioner, lotion, paper goods 4. **Cleaning supplies** — chemicals, cloths, gloves, liners Keeping the same order on every floor means muscle memory carries across the whole property. An attendant covering an unfamiliar floor still knows where everything is. ## 2. The eye-level rule The most frequently lifted items belong between knee and shoulder height. Reserve the lowest and highest shelves for bulk backstock and rarely-needed items. This reduces bending and reaching, which reduces fatigue and injury across a full shift. Terry and linen take the reachable band because they are lifted most. Chemicals sit at the bottom for a second reason that has nothing to do with reach: a container that leaks must never be positioned above clean linen. ## 3. What the survey found Eight closets were drawn by hand, one per floor, on 15 June 2026. Reading them against the property's own records — the layout notes, the property-walk record and the photo catalogue — settled the legend without anyone needing to be asked. The property has two towers: a tall one running floors 1–8 (264 rooms) and a low one stopping at floor 4 (136 rooms). **These closets are the tall tower's**, so the count that matters is how many rooms in *that* tower sit on each floor — not how many rooms the floor holds in total. | Floor | Rooms served by this closet | Racked walls | Unusable | |---|---:|---|---| | 1 | 13 | left + bottom | right wall | | 2 | 36 | right + bottom | bottom-right | | 3 | 36 | right, full height + bottom | bottom-right | | 4 | 36 | right + bottom | bottom-right | | 5 | 36 | right + bottom, short run | bottom-right | | 6 | 36 | right + bottom | bottom-right | | 7 | 36 | right + bottom, full | bottom-right | | 8 | 35 | right + bottom | bottom-right, heaviest | **That distinction is not pedantry, and getting it wrong produced a false finding in the first draft of this document.** The survey sheet recorded rooms per *floor* — 44, 71, 71, 71, 36, 36, 36, 35 — which is the whole building read across both towers. Attaching those to a single tower's closets made floor 3 look like it served twice what floor 7 served, and generated a puzzle about racking that does not track demand. It was an artifact. Floors 2 through 7 each serve **36 rooms from this tower**, and the puzzle disappeared when the numbers came from the room database rather than from another document. > **The rule underneath it: never let one derived document verify another.** The > survey and the worksheet agreed with each other and were both describing a > different denominator than the one the closets actually serve. Only the room > list settled it. **Three findings.** **Seven floors share one layout. Floor 1 does not, and it should not.** Floor 1 is not a supply closet, it is the central: the laundry plus the main storage room. It receives the chute rather than dropping into it, and it restocks directly rather than being restocked. Its shelving sits on the opposite wall because it is doing the opposite job, and it carries the fewest guest rooms of any floor in the tower. The outlier is not an inconsistency to correct — it confirms the model. **Every floor has a dead corner in the same place.** Bottom-right on all seven, which is systematic rather than seven separate accidents: electrical and fire panels carry a mandatory keep-clear envelope, and the property walk recorded the red keep-clear striping physically on the floor. Red on the drawing is red on the floor. **Run length varies across closets that serve identical demand.** Floors 2 through 7 each serve 36 rooms from this tower, yet the racking is visibly different: floor 3 runs full height, floor 5 is a short run, floor 7 is full. Same job, same room count, different shelf. This is the standardisation opportunity, and it is a cleaner one than the puzzle it replaced. When demand differs, unequal racking might be justified. When six closets serve the same 36 rooms, unequal racking is just history — whoever installed each one made a local decision, and nobody compared them afterwards. **Racking should be sized to rooms served; where it is not, that is a finding about the racking, not a reason to lower the par.** **What the fixed features give you.** The soiled chute and the service lift appear identically on eight of eight floors. Soiled drops, clean returns. Both are building features, which is why they are the two things a standard design can rely on. --- ## 4. The error this survey caught Par levels on floors 2–8 had been built on **323 rooms**. Those floors hold **356**. 323 was never a room count. It was the number of rows on a housekeeping status report dated 10 June 2026. A status report lists only the rooms that carried a status that morning — it is a snapshot of work, not a list of the building. The proof is that the report's floor-by-floor row counts were 66 / 67 / 58 / 30 / 34 / 35 / 33, exactly the numbers printed on the old worksheet. A work list had been read as an inventory. | Floor | Old sheet | Actual | Short by | |---|---:|---:|---:| | 2 | 66 | 71 | 5 | | 3 | 67 | 71 | 4 | | 4 | 58 | 71 | **13** | | 5 | 30 | 36 | 6 | | 6 | 34 | 36 | 2 | | 7 | 35 | 36 | 1 | | 8 | 33 | 35 | 2 | | **Total** | **323** | **356** | **33** | Every floor was under-provisioned; floor 4 was stocked as though it had 58 rooms when it has 71. **The room count was only half of it.** The old sheet also counted a queen room as one bed. A queen room here has two beds and therefore two sets of sheets, and 156 of the 356 rooms on those floors are queen rooms. The sheet was blind to 156 beds. What that produced on floor 2 was 21 sheet sets for 66 rooms — one set per three rooms — and 62 bath towels for 66 rooms. Those numbers were never going to stock a floor. **The rule that comes out of it: never build a par off a status report. Build it off the room list.** A status report answers "what happened this morning." A room list answers "what is the building." Only one of them is a denominator. --- ## 5. Setting par from the room list ``` Par = (rooms served x linen per room x change frequency) + safety stock ``` - **Rooms served** comes from the canonical room list, never from a work list. - **Linen per room** is bed-count dependent, not room-count dependent. Count beds. - **Change frequency** differs for checkout and stayover; keep them separate. - **Safety stock** is a buffer for spikes and late deliveries; 20–30% of the cycle quantity is a common starting point. Scale run length to rooms served, not to the space available. Where the racking cannot hold the computed par, that is a finding about the racking, and it should be recorded as one rather than absorbed by quietly lowering the par. --- ## 6. The five-day rollout One floor per day, starting with the floor whose par was most wrong. Work during the lull between the morning board and the afternoon arrivals so nothing stops. 1. **Day 1** — strip and re-rack the worst floor. Label to the four zones. Record what the racking cannot hold. 2. **Day 2–4** — repeat, one floor per day. Each floor is faster than the last because the standard is now familiar. 3. **Day 5** — walk every floor with the worksheet and confirm the zones match. Any floor that does not match is a training gap, not a design gap. Do not roll out to all floors at once. A standard that is half-installed across eight floors is worse than the inconsistency it replaced, because now nobody can predict which version they are walking into. --- ## 7. Carrying the method to a second property Everything above is a recovery. The par was wrong, the survey found it, the room list corrected it. That is the harder version of the job and it is the common one, because most properties already have a worksheet somebody trusts. The transfer case is not automatic, and this section is written as **preconditions you check against your own building** rather than as a description of anyone's. That is deliberate: version 2.0 described a second property's condition from outside, got it wrong, and the correction is recorded at the top of this document. **Check these before assuming the method carries:** - Does a canonical room list exist, and is it distinguishable from a work list? If the only room-level document is a status report, section 4 applies before anything else does. - Is the bed mix counted rather than inferred? Room count does not imply bed count. - Do the closets share a layout across floors, or does each need its own survey? **What transfers directly:** - The four zones and their order. Nothing about them is property-specific. - The eye-level rule and the chemicals-lowest rule. - Par built from the room list, with beds counted rather than rooms. **What has to be re-measured, because it does not transfer:** - **Room and bed mix.** A small property is not a scaled-down large one. The queen-room share drives linen par and has to be counted where the closets are, never inferred from Fairbanks. - **Closet geometry.** Fairbanks' standard exists because seven floors happened to share a layout and two building features were fixed on all eight. That is a finding about one building, not a law. A second property needs its own survey. - **Floors served per closet.** A small property may not warrant a closet per floor at all, and the right answer may be one central plus a cart route. **The one case where the method is cheaper than it was here:** If a property has not yet built a worksheet, the room list can be made the denominator from the first day and a status report never becomes a candidate. The rule from section 4 stops being a correction and becomes a constraint. That is a statement about sequence, not about any particular building — and whether it applies is answered by the preconditions above, by someone holding that property's own register. **The honest limit.** The Fairbanks work is a survey of one property with eight drawings, and the run-length question in section 3 is still open there. Nothing in this guide should be read as a benchmark. It is a method with one worked example and one property in front of it, and the second one will produce its own findings that this document does not anticipate. --- *Version 2.0. Superseded versions stay listed rather than disappearing. Figures are drawn from the 15 June 2026 hand survey; no guest or staff information appears in this document or in any drawing it contains.*