Effective Manufacturing Analytics Requires Cross-Functional Data Sharing

by   |   November 26, 2012 6:34 pm   |   0 Comments

BurrowsPortrait 240x300 Effective Manufacturing Analytics Requires Cross Functional Data Sharing

Bob Burrows

In many manufacturing companies, a process known as Sales & Operations Planning (S&OP) aims to tie together sales forecasts, manufacturing plans, raw material purchases and inventory management. The logic is simple: figure out what customers are likely to want to buy, make it, and then sell it to them when they knock on the door.

The problem? In almost as many manufacturing companies, executives will readily point to a litany of S&OP failures: poor forecasts, lack of coordination between sales and production, turf wars, excessive inventory and a lack of responsiveness. In short, goes the received wisdom among analysts, academics and industry insiders alike, S&OP is a good idea that is often let down by sloppy execution.

In a just-published book entitled The Market‑Driven Supply Chain, longtime supply chain consultant and manufacturing executive Bob Burrows looks at what goes wrong with manufacturers’ S&OP processes, and sketches out what he terms “the seven guiding principles of the design of a market‑savvy S&OP”.

The bottom line: In a successful S&OP process, good analytics has a critical role to play. And the other day, I sat down with Burrows to discover how.

Data Informed: So where do manufacturers go wrong with S&OP?

Related Stories

Supply chain leaders know they have a big data problem.

Read more»

Five supply chain opportunities for big data and predictive analytics.

Read more»

Integrating sustainability into corporate decision-making.

Read more»

Bob Burrows: They treat customers as a single homogenous group, with very little attempt to segment them by metrics that actually matter. For example, they segment them by factors such as size or geography, but not by how they use the product—which can have a significant impact on demand, and on demand predictability. In other words they collect an enormous amount of data, but fail to put it in context.

So how can analytics provide that context? What should companies do differently?

Burrows: You have to start with the fact that usually, there isn’t a clear-cut way of distributing data around the organization in a way that helps people to communicate what it means. The distribution function talks about SKUs and warehouses; manufacturing talks about production lines; sales people talk about territories and regions. There’s a lot of analysis going on, but it’s all within functions.

In the book, I describe how Goodyear used S&OP to enable manufacturing and the sales and marketing function to communicate more effectively. Goodyear had production lines that were very lean and efficient, having made extensive investments in techniques such as Six Sigma. But the result was that they were very efficiently making a lot of tires that no one was buying. Manufacturing would call up marketing after an S&OP meeting to ask: “Why are you not selling what we are making?”

Today, taking tires sold to Wal-Mart as an example, output is precisely tailored to the demand coming from individual Wal-Mart stores, which get deliveries once or twice a day via a nationwide network of specialist tire distributors, rather than Wal-Mart’s own distribution network. The production lines are slightly less efficient, but there’s virtually no finished goods inventory. It’s a saving of millions of dollars.

So how can manufacturers move forward to get to such insights? How do you start building an S&OP process that’s fit-for-purpose?

Burrows: You start by assembling a group of cross-functional people from each relevant part of the business, as any description of S&OP will tell you. But—and here’s the difference—they must look at the data using the language that the customer speaks, and not how they would look at that same data from within their individual functions.

Which group of customers is buying? What are they buying? Where are they buying it? Why are they buying it? Don’t forecast or analyze by brand, but instead use whatever imperatives drive the customer’s thought processes. In short, how does the customer think? And then map that thinking onto your own processes.

As an example, we found that in a medical device manufacturer producing instruments for operating theatres, analyzing and forecasting by type of operation gave a better result than analyzing and forecasting by type of device.

In the book, there’s a chapter entitled ‘Managing By Analytics’, that begins with a discussion of effective S&OP teams. Why that focus on team work, in a chapter on analytics?

Burrows: Because most S&OP teams aren’t proper teams, and aren’t effective. Teams work best when analytics rule the discussion. Yet we often find S&OP meetings being run by one department, or dominated by sales lore or historical conventions that go unchallenged, but which drive decision making almost unwittingly.

7 Characteristics of a Proper Analytic

In The Market‑Driven Supply Chain Bob Burrows cites these qualities for good analytics:

  1. Cross‑functional
  2. ‘Big picture’
  3. Relevant
  4. Understandable
  5. Provides perspective
  6. Passes the “so what” test
  7. Validated

As I say in the book, when a meeting relies on subjectivity and the opinions of the most articulate, when the less‑demonstrative participants are not encouraged to have a voice, and when bias and prejudices dominate, it is not a team meeting. The meeting is a power struggle, and the power of the team is lost. A strong team culture calls out these issues and insists on analytics.

In the book you talk about the ‘seven characteristics of a proper analytic’, in an S&OP context. Tell us more.

Burrows: Most people don’t have any idea of what a good analytic is. They are using statistics as a tool for describing data, and not as a tool for informing action. In other words, they aren’t trying to get the data to tell them a story about what will happen—they’re just describing what has happened in the past. Which isn’t the point. As a result, many analyses are simply a total waste of time in preparation and presentation, because they don’t tell a relevant story.

Freelance writer Malcolm Wheatley lives in Devon, England, and can be reached at editor@malcolmwheatley.co.uk.





2f4030c6 41e3 4507 9efb 8be1e3aad05d Effective Manufacturing Analytics Requires Cross Functional Data Sharing



Tags: , ,

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>