| Day | Actual (c/L) | Predicted (c/L) | Diff | Cost for your tanker | Recommendation |
|---|
Past rows: actual wholesale prices alongside what the model predicted for that day. Future rows: model's call for the next 5 business days. Anchored on 2026-06-10.
This week · 08 Jun · Melbourne operator, 1 petrol + 1 diesel tanker (30,000 L each)
| Fuel | Mon price (c/L) | Bought at (c/L) | Decision | Saved |
|---|---|---|---|---|
| Petrol | no decision data | |||
| Diesel | 191.20 | 191.20 | Buy Mon (no defer) | $0 |
| Total saved this week: | $0 | |||
Running ledger · 2 weeks of decisions tracked
| Week of | Saved (1 petrol + 1 diesel tanker) |
|---|---|
| 01 Jun 2026 | $1,620 |
| 08 Jun 2026 | $0 |
| TOTAL SO FAR | $1,620 |
How this works: Imagine a Melbourne operator who needs one petrol and one diesel tanker per week (30,000 L each). Baseline: they buy at Monday's price every week, no thinking required. Informed: they read this bulletin Monday morning and defer to the day the model said would be cheapest (using only forecasts available that morning, with no hindsight). Saved = baseline cost − informed cost. Negative numbers mean the model picked the wrong day that week. Scales linearly with your own tanker volume and frequency.