Margin as a live signal. Bottleneck as today’s answer, not next quarter’s.
Margin and bottleneck have lived in the finance and ops worlds as quarterly exercises. By the time the number arrives, the decision that produced it is two months old. Polymr turns both into live signals by reading the same records the other five workflows write to. Yesterday’s PO is on today’s scorecard. This week’s bottleneck moves on its own when the constraint moves. Margin and bottleneck stop being review meetings and start being morning glances.
Execution tracking is the live floor the margin layer reads from.
The execution lane carries every WO actual the margin layer uses. Throughput per work centre, scrap per lot, schedule adherence per shift. The view below is the live source; the margin and bottleneck numbers are calculated from it, not typed in.
Execute tracking · WO actuals
Shift A · 06:00 to 14:00 · 6 active WOs · last actual posted 09:38
| WO | Item · lot | Work centre | Progress | Scrap | Adherence |
|---|---|---|---|---|---|
| WO-1124 | PMR-4031 · L-A26-118 | WC-CR-01 · cure | 72% | 3.2% | on plan |
| WO-1126 | PMR-4031 · L-A26-122 | WC-MX-01 · mix | 48% | 0.4% | + 12m |
| WO-1131 | PMR-4406 · L-A26-119 | WC-PK-01 · pack | 91% | 0.1% | on plan |
| WO-1134 | PMR-HF-22 · L-A26-117 | WC-ML-02 · mill | 28% | 1.8% | + 44m |
| WO-1138 | PMR-CN-44 · L-A26-121 | WC-LA-01 · lathe | 64% | 0.6% | on plan |
| WO-1141 | PMR-SR-FK · L-A26-123 | WC-FN-02 · finish | 12% | 0.0% | queued |
The capacity heatmap shows the bottleneck before the planner asks.
Work centres run down the rows; days run across the columns. The Mix line goes red on Wednesday because three WOs all need the same window. The heatmap surfaces the binding constraint on its own; the planning screen rebalances against it without anyone opening a separate report.
Cost variance tied to the decision that produced it.
Standard cost was the answer last quarter. This week, the actual per-unit cost moves with every PO award, every yield event, every freight surcharge. The variance table below names the cause on the row, not in a footnote at the bottom of a slide.
| Category | Std / unit | Actual / unit | Var | Source |
|---|---|---|---|---|
| Material . brake hub | $38.10 | $37.26 | -$0.84 | V-244 award PMR-4406 |
| Material . bearings | $3.05 | $3.21 | +$0.16 | V-201 mix shift |
| Labour . cure line | $9.40 | $9.18 | -$0.22 | Cure-1 reroute |
| Yield . hub forging | $0.00 | $0.32 | +$0.32 | Lot L-A26-118 scrap 3.2% |
| Freight . inbound | $1.20 | $1.38 | +$0.18 | Carrier surcharge |
| Burden | $5.40 | $5.40 | 0 | standard |
| Net per unit | $57.15 | $56.75 | -$0.40 | 7d rolling, n=4,218 |
Price intelligence: are we paying market, or are we paying habit?
The price intel panel compares our actual landed cost against market percentiles and against alternate-vendor quotes. Drift shows up as a tag; the recommendation is a row, not a phone call from the category lead. Every signal links back to the record that produced it.
We are paying 5.2% over market median on PMR-4124 against the trailing 90 days. V-244 quoted within 1c of standard last week. Switching half the volume saves $86 per kilo unit at current run rate.
The longer Polymr runs, the more this workflow carries.
Margin and bottleneck is the slowest of the six to pay back in the first weeks because it needs the other five workflows to build up history first. Four to six weeks in, the history is deep enough that the numbers start matching what the controller would have produced by hand at quarter-end. After ten to twelve weeks, the numbers start surfacing things the controller would not have caught at all: small shifts in vendor mix, single-lot yield problems, week-over-week bottleneck changes that used to hide in the noise. This workflow is the case for staying with Polymr. The first five pay back in the same week. This one is the payback that keeps compounding.
- Wk 1-2Ledger pulls in. First attribution rows replace finance roll-ups.
- Wk 4-6Live variance band tightens. Bottleneck shifts surface within the day.
- Wk 10-12Single-lot yield anomalies caught pre-floor. Vendor-mix drift visible.
- Wk 20+The dashboard becomes the morning glance. Quarter-end roll-up redundant.
A multi-plant operation where vendor-mix drift was hiding 18% of category cost.
Anonymised engagement with an industrial-components operator. Looking at margin by vendor mix showed category-level drift that quarter-end summaries had been smoothing over for years.
- Situation
- Four plants procured overlapping SKUs from partially overlapping vendor lists. Each site ran its own planner and its own approved-vendor file. Central operations saw a roll-up only at month-end.
- What was breaking
- PMR-4124 bearing race was bought by three plants from three vendors at unit costs spanning 18%. Plant-level POs were sent before central could batch them. Vendor consolidation was a slide deck, never an action.
- Planning + purchasing
- Quote-to-procure
- Margin and bottleneck analysis
Where margin and bottleneck read from the rest of Polymr.
Vendor-mix attribution starts in the purchasing workflow. Yield attribution needs lot linkage from receiving. Industry context tells you where the margin gap usually hides in your shape of plant.
Supplier quotes, purchasing
Vendor-mix attribution starts here. Every award decision and every override is the row this workflow reads to attribute margin.
Receiving and delays
Yield attribution requires lot linkage from receiving. Without the receipt-to-lot graph this workflow produces aggregates, not attributions.
Industrial components
The industry view: multi-plant operations where the margin gap most often hides in vendor-mix drift.