How Do I Prioritize Orders in Production Planning?

What production planning tools do well and what is missing

Today is about mathematics, explained in a very unmathematical way, of course. After combinatorics, it's now scheduling's turn; the nerdy deep dives are still to come. The example: prioritizing orders. In another video we looked at whether I should accept an order, measured by contribution margin and on-time delivery. Today it's about the order sequence.

There are production planning tools, and we are always grateful when one is already in place. They can create a good plan with theoretical capacities in a theoretically ideal world, better or worse depending on the tool. Production planning or control tools usually cover the 80% we call planning quite well, and some can also adapt to operational events.

8:37 am: when the plan breaks

The crux lies in the remaining 20%. It's 8:37 am, the sales rep calls: very important customer, this order absolutely has to be squeezed in. If that only happened every few months, the planned system would work well and 98% would run according to plan. That would be a dream.

The reality: at 9:17 am the next sales rep calls. Nothing against sales reps; it is rightly their job to get the best for their customers. It just doesn't help if the customer gets a wrong delivery date or a promise that can't be kept. Everyone benefits from a good solution: plant management, executive management and sales. By noon there are already four such requests. Then an employee calls in sick, and on top of that a part has gone missing. That happens even in the best companies.

This is exactly where we come in with scheduling and rule-based prioritization. It's not about perfect planning where every employee and every machine is scheduled to the minute. It's about making good decisions in daily operations when orders are squeezed in, employees are sick or parts are missing. Circonomit is the everyday decision sparring partner, as a module next to the production planning system, the ERP, Excel or standalone. It doesn't have to be integrated into any system, but sensibly uses the planning data that already exists.

Defining rules for prioritization

Before planning perfectly and starting to over-engineer the master data again, we ask: which rules do we set? Family businesses in particular, with strong products, a strong market position and historically many variants, need to ask themselves this. A few examples, gathered from different customer projects:

• Are large customers more important than small ones?
• Is the custom-made product we have stood for for decades more important to us, or the clean series or small series with a possibly better contribution margin?
• Do we prefer to supply industries that are stable in times of crisis, such as infrastructure, energy or water, or the industry with the better margins?

For entrepreneurs it isn't always easy to decide. But we need rules so that in daily operations we can say: the sales rep called, it's a key account, we agreed together that it may be squeezed in. Or: this is a small customer from a priority 3 industry, we can do it, but only a week later.

Here lies the crux: being able to steer in daily operations, in line with the strategic goals. Strategy is useless if operations can't implement it. And operations are useless if they work completely past the strategy. Strategy must not happen behind closed doors and end up on a slide in a drawer while everyone in operations wonders what was actually decided. You can honestly ask yourself how well you have managed to operationalize the strategy. With simple, rule-based prioritization this translation works very well.

Scheduling as an optimization problem

Scheduling is the nice word for how I distribute orders according to these rules in a sequence across employees and machines. You can calculate scheduling in a very elaborate way using occupancy times and time series. Or you create value right away by defining the fewer than ten most important rules. After a few weeks you see in operations that on-time delivery, contribution margin, utilization or the satisfaction of customers in the core industry go up.

Quick nerd warning: every decision in a company is an optimization problem. You want to maximize or minimize something, there are things you can change and things you can't, in nerd terms variables and constants. Some things you can only change within a certain range; those are the limits of the optimization.

What excites us is finding super pragmatic solutions that make companies in Europe able to steer today, with our nerdy maths heart in the background. For the production manager, the result is the answer to how to react well when chaos breaks out in operations again. And executive management becomes more satisfied because the key KPIs and the long-term direction of the company fit together.