Analyzing products, demand, margins, and routings

Most job shop managers operate multiple businesses under one roof. Such high-mix, low-volume (HMLV) operations use the same business strategy, the same suppliers, the same facility layout, the same production scheduling, and the same equipment, all for what amounts to several different businesses.

Unlike OEMs and top-tier suppliers--which generally produce product lines in a relatively low-mix, high-volume environment--job shops struggle to develop rules and guidelines to rationalize their business and standardize processes, materials, policies, tools, routings, setups, and so forth--all completely understandable, given their diverse product mix. For this reason, they send disparate products--some high volume, some low volume, some repeat, some one-offs--through the same manufacturing processes.

A typical lean implementation in a job shop starts with selecting a few products produced per a long-term agreement. Takt times are calculated, one-piece-flow cells are designed, kanban lot sizes are determined--but then what? It’s true that a job shop can implement flow production for a portion of its product mix, but what about the rest? This is why product mix segmentation is so important.

A typical job shop--the polar opposite of the Toyotas of the world--manufactures a diverse mix of products (P) with different routings (R) produced in various quantities (Q) with varying revenue and profits ($) and demand repeatability (T). These parameters for each different product make up a PQR$T Analysis, a comprehensive data analysis tool that job shops can use to do a product mix segmentation. This ultimately helps the job shop identify the various “businesses” it is knowingly (or maybe unknowingly) operating, from somewhat predictable repeat work to fabricated prototypes.

The PQR$T Analysis separates products that produce high revenue or (ideally) profits from those that do not--that is, high-value jobs are treated differently than low-value jobs. It also breaks up the product mix per their routings – that is, those that belong to a product family and those that do not.

Defining Product Mix

Job shops can process hundreds or even thousands of different products a year, and classifying them can be a monumental challenge. Here, the PQR$T Analysis can help. Specifically it relates a product’s quantity and demand repeatability (Q and T) by defining runners, repeaters, and strangers.

A runner is a product or product family that has sufficient volume for a dedicated facility or manufacturing cell. This doesn’t mean that such facilities or cells need to be used continuously--few job shops manufacture such volumes. Instead, these runners may justify a cell or facility that’s used as needed and, most important, is not shared with any other product or product family.

A repeater represents repeat work, but volumes aren’t high enough to dedicate a cell or facility to produce them. Repeaters are scheduled for production in regular intervals. For instance, toolroom personnel may know that a certain die is needed each Tuesday morning. Suppliers get used to that regular order, setup resources are made ready, forklifts may be standing by, and so on.

A stranger is a product or product family with low or intermittent volumes. These are fitted into the schedule around the regular slots dedicated to repeaters.

From all this, the final PQR$T Analysis creates a multidimensional data plot that can reveal much about a high-mix, low-volume operation. Volume/quantity and demand repeatability (Q&T) are analyzed for runners, repeaters, and strangers. Products are also grouped into those producing high or low profit margin (or revenue). Finally, routings (R) are analyzed for both products that are part of a product family and those that are not.

Benefits of PQR$T Analysis

Why bother with such an analysis? It is because it can uncover inefficiencies and identify the core strengths of an operation. If a job shop manager knows exactly what makes the most money and what doesn’t, and why, he can use this for immediate improvements as well as long-term planning.

When it comes to immediate improvements, the analysis can help managers set up flexible fabrication cells for the strangers in the product mix, with flexible, quick-setup machines. Another cell (or even a separate area) could be dedicated to produce spares or repair items, while another could be dedicated to prototype work. Having a dedicated prototype area would help managers identify and analyze emerging and new business that might eventually end up in the production area.

The analysis also can aid business negotiations. For instance, one job shop may not be able to make high margins on certain work because the work is outside the shop’s core competency. After analysis, a manager would know that these strangers--or even runners--aren’t helping the company grow. But that doesn’t mean these jobs wouldn’t benefit other job shops in the area. Why not swap these jobs with these other shops that may be able to benefit from work that fits their core competence, workforce skills, manufacturing technology, and supplier capabilities? It’s about competing through collaboration. If a manager has concrete data revealing what makes money for the company and what doesn’t, he can negotiate with competitors and swap jobs with similar revenue. What may be unprofitable for one shop may be very profitable for another, depending on the equipment and processes each job shop has on their floor. Of course, a manager could end up with a bad deal unless he has the necessary intelligence, and here is where PQR$T Analysis comes into play. It lets the business data do the talking, instead of relying on well-worn strategies like waste elimination that have run their course.

Beyond PQ Analysis

For years, low-mix high-volume operations have utilized PQ Analysis, which segments the product mix (P) based only on the volumes or quantities (Q) in which they are produced. But despite its popularity in these production environments best represented by the likes of Toyota, Honda, John Deere, or even Whirlpool, the PQ Analysis may not be suitable for the job shop, simply because it does not consider the R (routings), the profit ($), and demand repeatability (T) of the different products.

Over the next few installments of this column, we will build on the different analyses (PQ Analysis, PQR Analysis, PQ$ Analysis, and finally PQ$T Analysis) that, when combined, make up the PQR$T Analysis that has proven successful for various job shops pursuing lean. Each analysis will progressively reveal more about the reality of a job shop’s current state, which of course is the ultimate goal. After all, the more job shop managers know, the better they can optimize their operations.

For other articles in this series, see the following:

Analyzing product mix and volumes: Useful but insufficient

Minding your P’s, Q’s, R’s--and revenue too

PQ$ Analysis: Why revenue matters in product mix segmentation

Why a timeline analysis of order history matters