Most operators can't tell you which of their four hundred SKUs actually make money — only which ones sell. Allometry maps the whole line against realized margin and real demand, so you can weed the dead tail, reprice the volume traps, and scale what earns.
So the line only grows. Every dead SKU still consumes a slot in the price book, a row in inventory, a line in training, and a decision from a rep at quote time. The cost is real and never on anyone's P&L — until you can see it per SKU.
Margin and demand are independent axes, and most portfolio reviews collapse them into one revenue ranking. Separate them and the action becomes obvious per SKU.
The Pareto shape is always worse than people expect. A handful of SKUs carry the gross profit, a middle band pays its way, and a long tail earns nothing while consuming inventory, training, and quote-time attention.
Cutting the tail is defence. The offence is recognising that what you already deliver can be packaged as a product, a service, an outcome, or a digital good — each with a different margin. You don't need a new capability. You need a new form.
Margin rises as you move up the stack because you're selling less stuff and more certainty. The catch: you can only sell an outcome if you can measure the cost of delivering it — which is exactly what the four loops produce. In the illustrative 412-SKU book below, shifting 20% of mix up one tier moves blended margin +6.4 points.
Once delivery is structured, the adjacent revenue form is visible. The pattern holds well outside our vertical — which is why the same backbone runs other operating models.
SKU Morphology reads every stage and feeds back into what you offer next. Expansion returns to the start — the loop closes rather than ending.
The spine doesn't care whether the unit is a charger, a cover, or a seat. Where the primary object changes, the loops stay — which is how the same platform runs very different businesses.
The primary object is a cover, a room, or a booked hour. Occupancy is the meter, cost-per-cover is the margin, and the adjacent possible is retail and licensing.
Location → Seat → PulseThe primary object is a case at a door. Velocity per SKU per retailer replaces jobs per address, and the adjacent possible is category data.
Door → SKU → PulseA containerised agentic version where the object is an account and a seat. Same swarm, same veto, same audit trail — no physical asset required.
Account → Seat → PulseA price book and twelve months of closed jobs is enough. We'll map every SKU on margin and demand, quantify what the dead tail costs you to keep, and propose the net-new offers hiding in what you already deliver.