Aligning the supply chain with business strategy · Lesson 1

Customer promise and the forecast–order boundary

4 min read

Lesson objectives

Apply Customer promise and the forecast–order boundary to a bounded case and justify the result and its limits.

The customer promise here is the time from accepting an order to delivery. 'High service' alone cannot determine a fulfilment method: first ask what the customer will wait for, what can be stocked economically, and how long each operation takes after an order arrives. The customer-order decoupling point separates work driven by a forecast from work triggered by a confirmed order. Stocking finished goods permits dispatch soon after the order but exposes the firm to unwanted colours or configurations. Stocking common components and assembling the chosen configuration avoids many finished-good variants but adds assembly time after the order. Starting manufacture after the order requires the customer's tolerance to cover fabrication, assembly, inspection and transport. This is a decision about where inventory and capacity sit, not a change of label.

Compare the promise with the critical path after an order. If transport takes one day and assembly takes two, assemble-to-order cannot meet a two-day promise without a real scheduling or transport change. Fast dispatch capability does not automatically make finished-goods stocking preferable: count variants and obsolescence exposure, then check whether common components can serve several configurations. One firm can use different responses for different segments. The defensible choice follows a stated promise, demand pattern and constraints; revisit it when lead time, capacity or mix variability changes. A stable total forecast does not guarantee availability of every colour: the average can conceal a changing mix.

Also separate an order qualifier from an order winner for each segment. Conformance to a lamp specification may be a condition for consideration, while fast delivery wins among conforming offers. Sacrificing required testing to improve apparent lead time removes the qualifier; speed does not compensate for that loss. Begin a decision table with the promise, what is stocked before the order, the operations still required, and slack. Compute slack as the promise minus the elapsed time of remaining operations; neither variant count nor average demand is a number of days to subtract. Positive slack is not a delivery guarantee: stock may be reserved, capacity may have a queue, or the assumed transport slot may be unavailable. Attach an operational check to the choice: unreserved available finished stock for Standard, and confirmed components and assembly capacity for Custom. If the check fails, the real options are to secure feasible stock or capacity within acceptable cost, or renegotiate the promise with sales. Entering an earlier due date in a planning system does not shorten the physical path.

Example

Hypothetical training scenario in Riyadh: a lamp maker sells two segments. Standard is one product with weekly demand of 80–100 units and a two-day delivery promise. Custom uses a shared housing and three shade colours; demand by colour is irregular and customers accept five days. Packing and transport take one day, assembly and testing take two, and making the housing adds three more days. For Standard, finished-goods stock leaves one day after the order, meeting the two-day promise with one day of slack if inventory is available. For Custom, stocking housings and shades then assembling to order leaves 2 + 1 = 3 days, meeting five days with two days of slack while postponing the colour decision. Starting housing manufacture after a Custom order takes 3 + 2 + 1 = 6 days and misses the promise. The result is two responses for two segments; component availability and assembly capacity must still be checked before committing. These numbers are fictional and make no claim about the Riyadh market.

Exercise

Produce a decision table with a justified choice for each segment: in a new training case, a single standard device has a two-day promise and available finished stock; packing and transport take one day. A custom device has four configurations, shared parts and covers are available, assembly and test take two days, and transport one; its customer accepts four days. Manufacturing parts after the order would add four days. Calculate each post-order path, locate the stock for each segment, and name one capacity constraint that could invalidate the promise despite correct arithmetic. Required output: a decision table with justified choices, each path's time and its limit.

Check your answer

Exercise solution: inputs are the Standard promise of two days, Custom promise of four, one day to ship, two days to assemble Custom, and four days to make its parts after an order. Decision table: Standard—hold finished stock; post-order path one day; slack 2 − 1 = one day, conditional on an unreserved unit. Custom—hold parts and covers, assemble after order; path 2 + 1 = three days; slack 4 − 3 = one day. This delays configuration commitment and meets the promise if assembly capacity is available. Making parts after a Custom order takes 4 + 2 + 1 = seven days and misses the promise by three, so reject it. Check that the table measures work after the order rather than manufacture already completed for stocked parts, and names the inventory position. If the assembly line has a two-day queue, the Custom path becomes 2 waiting + 2 assembly + 1 transport = five days; the four-day promise fails unless scheduling is changed and confirmed or the promise is revised. The case does not establish an optimal stock quantity or stockout probability because demand distributions and capacity are not supplied.

Lesson summary

With stocked parts, Custom takes two assembly days plus one transport day, three in total and one day of slack against four; a two-day assembly queue raises it to five and misses the promise by one.

Report an error in this lesson

Lesson objectives

Lesson summary

Share this page