Multi-Site Catering Procurement & Service Optimisation · Data Sheet
Catering Procurement Financial Results
Quantified savings, service outcomes and result-status architecture from a national multi-site catering review across nine provinces.
Publication basis
Anonymised, percentage-led and status-separated. Finalised and contract-supported workstreams are reported; identified opportunities and continuing negotiations are excluded from the attained result rather than blended into it.
1 · Result classification
Each workstream and the status of its result
Scroll table sideways →
| Workstream | Status | Result |
|---|---|---|
| High-volume Gauteng anchor site | Negotiated / attained | Per-meal economics reduced 8.5% across a recorded annual volume of approximately 28,900 meals. |
| Two additional Gauteng sites | Negotiated / contract-supported | A common regional route reduced the two legacy rates by approximately 23.8% and 32.7% — a combined 25.3% against their reconstructed baseline. |
| KwaZulu-Natal three-site cluster | Negotiated / contract-supported | Incumbent rate reduced 13.3% after operational menu redesign, with quality and quantity requirements preserved and the service formalised. |
| Combined finalised workstreams | Finalised result architecture | Approximately 14.6% annualised reduction on the addressed spend base; approximately 6.6% of the reconstructed national baseline. |
| Additional provincial opportunities | Identified, not included | A further material opportunity was modelled in one province and negotiations continued in others. Excluded from the attained result. |
2 · Baseline scale
The evidence population, and what each part established
Scroll table sideways →
| Measure | Recorded scale | Result relevance |
|---|---|---|
| Transaction population | 3,672 populated rows | Created the spend trail used to identify supplier, branch, invoice and project-code patterns. |
| Supplier / payment-channel entries | 66 entries, 62 with non-zero spend | Demonstrated the fragmentation, and the need to tell formal caterers from ad hoc purchasing routes. |
| Operating locations | 26 locations across 9 provinces | Required geographic segmentation rather than assuming one national caterer could serve every site. |
| Project / cost codes | 33 codes | Supported branch-level reconciliation and exposed changes in location and operating structure. |
| Meal volume | Approximately 127,000 meals | Provided the volume leverage used for negotiation and regional sourcing design. |
3 · Workstream economics
Indexed baseline against future state, with the service effect
Scroll table sideways →
| Workstream | Baseline | Future state | Reduction | Service effect |
|---|---|---|---|---|
| Gauteng anchor site | 100.0 | 91.5 | 8.5% | Existing provider retained; rate renegotiated against high recurring volume. |
| Gauteng migrated site A | 100.0 | 76.2 | 23.8% | Regional route, broader buffet specification and standard beverage. |
| Gauteng migrated site B | 100.0 | 67.3 | 32.7% | Same regional route; menu breadth and service model improved. |
| Two migrated sites combined | 100.0 | 74.7 | 25.3% | A common regional supplier architecture created scale without a blanket national model. |
The service-effect column is not decoration. A per-meal rate can always be lowered by serving less; these reductions were achieved while the specification held or improved.
Denominators
Why both 14.6% and 6.6% are published
14.6% measures the reduction against the spend the finalised workstreams actually addressed. 6.6% measures that same benefit against the full reconstructed national catering baseline.
Neither is more honest than the other; they answer different questions. The first describes how well the intervention performed where it was applied, the second describes what the organisation has secured so far across its whole estate. Publishing only one would mislead in one direction or the other.
Interpretation
These percentages reflect a specific fragmented catering environment with high branch dispersion and inconsistent historical purchasing. They are not a catering benchmark.
The transferable finding is methodological: in a dispersed service category, accounting data identifies who was paid but not reliably what was delivered, and the normalisation step against specification is what separates a real saving from a cheaper meal.