Why ETO and CTO manufacturers still plan in spreadsheets, and what to plan instead

Part 1 of 3: Planning what you haven’t designed yet

Put an S&OP manager from an engineer-to-order (ETO) or configure-to-order (CTO) manufacturer next to one from a make-to-stock company, and the difference shows quickly. For the ETO or CTO business, most of what will ship next quarter hasn’t been configured yet, and some of it hasn’t been engineered yet. Still, the plant needs capacity, suppliers need purchase orders and customers expect a delivery date they can count on. So why do so many of these manufacturers still plan in spreadsheets? Because they are trying to plan in a system built for make-to-stock products, retrofitted to accommodate ETO and CTO businesses.

Why the right system is hard to find

ETO and CTO planning doesn’t belong to one industry. It shows up in configured servers and test equipment, imaging systems and surgical equipment, irrigation systems and packaging machinery, forklifts and specialty trucks, aircraft interiors and ground support equipment. The products look nothing alike, but the planning problem is the same: a generic model, a set of options, some engineered content, and demand that arrives as projects and quotes. Because the problem is spread across so many industries, most planning software treats it as an exception, retrofitted onto a make-to-stock design rather than built for it.

Foundational problem of CTO/ETO and make-to-stock systems

Most planning systems, including the planning modules inside ERP, assume every product has a part number and a fixed bill of materials. Forecast the part, explode the BOM, check capacity, buy the materials. A heavy forklift with just four configuration choices (three engines, two cabins, three chassis and four option packs) already has 72 possible products, and customer-specific engineering makes the count unlimited. Most configurations sell once or never, so planning by part number is not an option. The result is ETO and CTO planning forced into a system built for make-to-stock, with spreadsheets filling the gap. They tell you to fix your data, but you really need a different kind of system. 

The cost of not leaving behind Make to Stock based systems

The impact on the P&L is significant. Bottlenecks that surface late cost capacity at the constraint, so backlog slips and revenue moves out. Quotes built on standard lead times hurt on-time delivery. Buffers sized by feel inflate inventory, while an options mix planned poorly or parts orphaned by engineering changes, turn into excess and obsolete stock. Some resources run overloaded while others sit idle, due to the inability to handle option mix impact on resource load.  Margin erodes through expedited freight, overtime and premium subcontracting to recover late orders. Planners and expediters spend their weeks reconciling spreadsheets and fighting fires, while systems costs are high from customizing the system.

Plan the model and the mix, not the part number

The answer starts with a different question. Instead of asking “how many of each part will we sell?”, ask “how many of each model will we sell, and what mix of options will they carry?”

Attribute-based planning makes that possible. Each configuration choice becomes an attribute of a generic model instead of a separate part number. Demand planning forecasts the model and uses attach rates to break it into options and components, by region, by customer or for a single large project. Inventory planning places buffers at the shared components that forecast uses, sized by their volume and variability. Supply planning turns the same mix into capacity and material requirements, that shifts as mix changes without a part number for every configuration.

One system from S&OP to S&OE

S&OP uses the model and configuration forecast to commit what can’t wait for an order: capacity, long-lead materials, sub-assemblies and strategic buffers. Sales and operations execution (S&OE) then plans the blend of configured orders and the forecast not yet ordered. The forecast reserves capacity and materials; the orders execute. Because both run in one system on one model, there is no hand-off to reconcile, and every order sharpens the forecast mix. That is the foundation Eyelit SIOP from Eyelit Technologies is built on.

What changes on a Monday morning

Sales flags a large fleet opportunity with a state roadworks department in the CRM system. The Northeast region is trending ahead of forecast. In a system built for make-to-stock, neither signal has anywhere to go but a spreadsheet. In a system built for ETO/CTO using attributes, the shift in the customer demand patterns uses the configuration predictions to update the demand for the option mix, and the planner sees the effect on capacity and long-lead time materials immediately.

Part 2 of the 3 part Blog series looks at the demand side: how to forecast the models and their option mix, and how to decide what to buffer and how much.