Corporate & Business · NPO & Social Impact · Procurement & Sourcing
The same saving reads as 15.9% or as 4%, depending on what sits under the line
Distributed buying hides inconsistent descriptions, pack sizes, once-off purchases and unrelated costs in the same account. Correct those first or a supplier comparison produces a precise-looking and unreliable percentage.
- 15.94%
- tender basket reduction, validated
- 9.5–23.8%
- range across the six locations
- 68
- tender lines defining the basket
- 4 → 3
- suppliers quoted, then shortlisted
Proof context: Distributed branch purchasing across six operating locations
The situation
Six locations buying the same things differently
Distributed stationery buying hides inconsistent product descriptions, pack sizes, once-off purchases, local habits and unrelated costs sitting in the same account. Each of those is individually small and collectively fatal to a comparison.
If they are not corrected first, a supplier benchmark produces a precise-looking and unreliable saving percentage. So the work was sequenced deliberately: rebuild the recurring basket, reconcile it, establish product equivalence, and only then let suppliers price it.
What KuTh did
Clean the basket, then compete it
- Rebuilt the branch spend baseline. Across six locations, separating recurring stationery from storage and other costs that had accumulated in the same supplier relationship.
- Reconciled the detail. Descriptions, quantities, pack sizes, monthly and annual treatment, and like-for-like substitutions — the work that makes two prices comparable at all.
- Built a defined tender basket. 68 product-and-quantity lines, giving every supplier a common basis to price rather than a request for a discount off their own list.
- Addressed what happens after award. Rate duration, escalation and commercial terms, because a tender price that is not held is a tender price that erodes.
Results
The same work, measured four ways
Scroll table sideways →
| Measure | Result | Status |
|---|---|---|
| Formal Phase 2 tender comparison | 15.94% lower basket cost | Validated / quoted — supplier agreements were still being formalised. |
| Location-level range | 9.54% to 23.82% | Validated / modelled. Different purchasing mixes produced materially different outcomes. |
| Corrected Phase 1 stationery-only view | Approximately 11% | Identified / modelled, before the Phase 2 tender comparison. |
| Combined denominator including storage | Approximately 4% | Context only — see below. |
Why the same saving can be read as 15.9% or 4%
The 4% figure is the same commercial outcome measured against a denominator that still contains non-stationery storage charges. Those charges were never part of the stationery sourcing exercise, and leaving them in dilutes the percentage by roughly three quarters.
The pack publishes it as context rather than as a result. It is included here for the same reason: it is the clearest available illustration of how a number can be technically accurate and still mislead, depending entirely on what sits underneath the line.
What is not claimed
No realised cash saving is claimed. The source records do not contain enough post-award purchasing data to establish one, so the result stops at a validated tender comparison with supplier agreements still being formalised.
A separate 15% discount was negotiated with an adjacent records and storage supplier. It applies only to specified qualifying lines, is conditional, and is deliberately not counted as a stationery result.
Two things flagged, not resolved
Found in the incumbent statements
Forty-eight repayment entries appeared in the incumbent's statements without the source records explaining what caused them. They were flagged for reconciliation rather than converted into a recovery claim.
Duplicate invoice patterns were also identified. Both sit in the published record as open items, because an unexplained credit is not evidence of a refund due and the engagement evidence does not take them further.
Supporting documents