December 1996 € Volume 6 € Number 12

Fast and Sure Response:
A Commonsense Approach

By Philip E. Quigley, CPIM

Speed, fast response, efficient customer response&emdash;words and phrases that drive materials professionals today. Everyone wants to be able to move faster, to stay ahead of customer demand and do it with less inventory. Some companies have done well and reaped substantial rewards&emdash;Wal-Mart and Toys R Us. Some have failed:

• Packard Bell misread the computer market and had plenty of 75MHz machines in stock but not the 120MHz or 133MHz machines the market was demanding.
• Apple Computer made the same type of mistake that led to its massive write off.
• GM had excess cars in inventory in several model lines this past summer.

The lesson is clear: If you don't move fast, you will pay for it in excess inventory, lost market share or both. But to move with speed and certainty takes careful planning and hard work. The planning side of the equation takes a carefully thought-out system (or process) that combines information systems and management action. The steps in the process are shown in Figure 1.

Figure 1 shows a five-step process, but note the two driving functions that affect the steps: Management Intent and Time. Management Intent shows the impact and requirement for the system to support management's strategy or purpose in the marketplace.

• Dell Computer has very carefully followed a strategy of direct selling to the PC customer. It has an extensive and sophisticated system to receive orders directly on the phone, configure the system and ship it within 24 to 48 hours.
• Caterpillar has designed systems to support its pledge to deliver any spare part required to get a piece of equipment operational within 24 hours on a worldwide basis.

Two very different missions with two very different systems, but one thing in common&emdash;management has defined very clear objectives for how the systems are to support the business strategy. Everyone in the organizations knows what the systems are to do and how their success will be measured.

The other message here is that management intent can change. Strategies and objectives change, so systems may have to change also.

For example, Caterpillar is now working on developing systems to allow it to monitor its products in the field and identify a problem while the machine is still operational. The system will then order and deliver a part before the machine goes down. The result is zero down time&emdash;an ambitious project that builds on Caterpillar's previous success. But it means major changes to existing systems, new systems and new ways of acting for Caterpillar.

Time is the second driver, and it depends totally on management intent and the marketplace. If you are supplying product to Wal-Mart or Compaq, you will have to respond on a daily or weekly basis. If you are supplying to Boeing or Caterpillar, responding on a weekly or monthly cycle may be adequate. System designers must understand the requirements of the marketplace and design their systems accordingly. Failure to be sensitive to the time issue can doom a new system or company.

Once the drivers of Management Intent and Time are understood, you can then move through the process of designing the system.


Step One: Collect the data
Obviously, collecting order entry data as quickly and accurately as you can is the first, but not the only, step. This means a point-of-sale (POS) system working on a real-time or daily basis in most cases. But now some other complicating factors come in. In today's com-petitive world just having an SKU number, quantity and date sold isn't enough. If possible, data should be collected on the buyer, the more the better. This means the POS system may have to interface with a payment system of some sort, such as a credit card. A good example of this can be seen in retail. Now retail customers at supermarkets can pay with credit cards, can use store cards to get discounts on products or earn "bonus" dollars for free gifts. The retailer and the manufacturer can collect a lot more data on the buyer and on his or her buying habits or preferences. But obviously, a lot more system and administration have to be developed.

Collecting data now also means collecting service, warranty and repair data for quality and sales information. This data is a gold mine for companies that know how to use it.

Example: Lexus has an outstanding sales and support system. Besides tracking all new car sales, it tracks warranty and service for each car on a nationwide basis. The system also collects sales leads from responses to magazine, TV and radio ads on a nationwide basis. Lexus can now identify and fix a problem before it has any impact on customers, and can proactively manage its customers to continue their buying of Lexus products and services, and more important, be a word of mouth sales representative for Lexus.

Example: Wal-Mart's system of order entry and tracking inventory is an example of a system that set a new standard in operations for the retail industry. This system is now directly affecting Wal-Mart's suppliers. But the system has grown from just an order entry system into a system that helps Wal-Mart and its suppliers to better manage customer relations and new product development.

Collecting the data is the foundation for all other steps and it looks deceptively simple, but anyone who has worked with systems knows how complex and costly this step is. Fundamental questions of amount and type of data, speed of collection, and security will have to be answered based on real business needs. Wrong answers to these questions can and have cost companies millions of dollars in cost, excess inventory, lost sales and market share.


Step Two: Analyzing the data
This step is where the data is reviewed and sense is made out of it. The review should tell you trends&emdash;where a product is strong or weak. The data should immediately tell you whether to produce more of a product or cut back production. The data could be telling you to cut prices, add features, do more advertising or even stop selling a product or come out with a new one. In other words, the data has tactical and strategic ramifications. Your first priority is to identify short-term tactical considerations then move to strategic.

Example: A computer company is analyzing its current sales versus the plan. It's seeing a shift from lower-cost units to higher-cost units with faster processors and more multimedia options. The immediate tactical decision is to switch production to the higher-priced units. The strategic decision is to push forward a new design that incorporates some new multimedia features versus a low cost simple unit for the home. This is an example of how one set of data has several uses.

Systems today will have executive reporting modules that can take raw data and graph and chart it in dozens of different ways. The key is to find the right format for the data and have a process ensuring that the right people will be looking at the data on a routine basis. The people reviewing and analyzing the data should have the background&emdash;in both intellect and experience&emdash;to understand what they are looking at. The lowest level possible should be doing the analysis&emdash;headquarters types can't be doing this. The analysis should include action to be taken. This recommendation should list different alternatives.


Step Three: Make a decision
After the data has been collected, analyzed and alternative actions identified, a decision has to be made. The key is having the right level make the decision in the appropriate time. The lowest level person should be making the decision, and be responsible for the results. Passing decisions up the chain of command only postpones the decision and can make it obsolete. But where and how the decision maker is identified and empowered is a senior management decision made at the design of this process&emdash;it cannot be a spur-of-the-moment decision.

Example: Shifting production from one model of computer to another is fairly straightforward and should be done by a production planner or master scheduler for the assigned product line. But there are problems&emdash;some real and some political. The real problem might be that management mistook the marketplace and demand. Reality is that adjustments must be made in the production plan, but changes may also be necessary in the advertising budget, in engineering efforts for the follow up on products, and so on. Now the political snake raises its head. How often have you seen necessary action postponed while everyone fought over who is to blame? The decision-making process must be designed to quickly get decisions made and action taken. Senior management must take careful and prolonged action to ensure the analysis is clear, that decisions are made quickly and that lessons are learned quickly and completely in an environment that minimizes finger pointing.

Step Four: Implement the decision
The decision has been made &emdash;now all of the thousands of follow up actions must be taken quickly and competently. The system must allow replanning to be done on a reasonable basis. Things to think about include:

• If you have an MRP system, can you run it daily and generate the changed requirements?
• What if you have to play what-if and look at several different scenarios? Can you do it quickly and efficiently?
• In today's world, fast response means your suppliers must react quickly. Does your system tie them in effectively?
• If you need to tie in other groups&emdash;engineering, sales, advertising, accounting, transportation&emdash;does your system allow you to integrate their activities?
• Beyond information systems, do you have the people in place who can and will react quickly and competently?

Interesting points aren't they? How many times have you seen decisions made that fall apart once implementation begins? How many times have suppliers built to old schedules because no one told them? Engineering releases the wrong prints? Or when everybody in manufacturing and engineering performs but no one told sales, so the advertising campaign calls out the wrong product?

The same amount of time and energy spent designing the information flow and decision-making process must be spent in ensuring the systems and processes that execute the decision work effectively and efficiently.


Step Five: Learning
Now that information has been collected and analyzed, decisions made and implemented, it is time to sit back and think through what the organization has learned. This is a key step because, if done properly, it can set the stage for further action in the organization.

Example: Let's look at the computer company that misread the demand for particular machines. Further analysis can be done on why the original forecast was wrong. Which assumptions didn't prove out and what does it mean? New advertising could be developed, new distributors added, changes in design could be made, or new forecasting techniques could be developed. In other words several things could be done&emdash;the key is learning and taking action. The marketplace is dynamic. You and your organization must be dynamic and be willing to learn.

The steps in this process are simple and straightforward but the simple things are the hardest to do well on a consistent basis. Time, energy and money must be spent on doing these things well. But remember what happens when you don't do them well&emdash;you have inventory you can't sell and lost sales because you didn't have the product that the market wanted.