By many measures production at Dunlop Tire Corp. of South Africa was rolling right along. Factory volume was up, waste was down, and employee productivity routinely exceeded incentive levels. But beneath those surface metrics, the company was awash in inventory, overburdened with slow-movers and short of best-sellers - a classic case of "push" production being out of sync with the market's demand "pull."

Dunlop's solution was to implement a new forecasting process based on Inventory Optimization software from Ability, an initiative that fueled a significant boost in market share while producing millions of dollars in inventory reductions. Ability is a South African company based in Johannesburg.

"When we first began this project, we knew we were overstocked, but the numbers that Ability showed us were embarrassing," says Leigh Harris, director of logistics for the South African operation. The natural tendency is to disregard or be skeptical of such numbers, especially when they are presented by someone trying to sell a product, he says, but in this case rapid improvements gave proof to Ability's initial evaluation. "Once we got a few months into the implementation, we got down to figures so close to Ability's projections that there were proved right," he says.

Inventory levels for finished goods fell 37 percent within the first six months, prompting the company to extend the application to other aspects of the operation. Results in those business areas also exceeded expectations; Dunlop experienced a 24 percent reduction in raw materials stock on hand, a 22 percent reduction in excess raw materials stock, a 48 percent reduction in potential stock-outs in raw materials supplies, and a 39 percent reduction in potential stock-outs for tools and spare parts, the latter being a condition often addressed by air freight shipping or emergency buying measures. Add to that a 4 percent boost in local market share, and it's easy to see why Harris and his fellow managers smile a lot these days.

"The tire market in South Africa is largely generic, as all the manufacturers are making roughly the same sort of product to roughly the same sort of standard," Harris says. "Consequently, the only time you can beat the other guys is when you have the right tires on the dealer's shelf when consumers go shopping. With the help of Ability, we now have the right tires in the right places more often than our competitors, and that gives us a pronounced edge in this highly competitive market."

A well-established South African business headquartered in Johannesburg, Dunlop began operations there shortly after World War II. During the apartheid years, the South African operation formally was separated from Dunlop's North American business. The company produces a wide range of tires and it imports tires and tubes as well, mostly from Korea and Taiwan.

"We manufacture a range of approximately 1,000 different stock lines, from very large tires used for mining and earth moving to passenger tires to small tires used for all-terrain vehicles and motorcycles," explains Harris. Dunlop distributes its tires domestically via a network of branch operations in major South African cities, which in turn feed tires to unaffiliated commercial tire retailers where consumers shop. The company also exports to Europe and Scandinavia a large percentage of automotive passenger tires produced in South Africa and sells truck, bus and earth-moving equipment tires to the U.S., South America, Europe and, on a much smaller scale, to the Far East.

To feed its factory operations in Johannesburg, the company imports rubber from Asia; steel from a range of suppliers; and a variety of chemicals, mostly from Europe.

By the mid 1990s, a considerable amount of money was tied up in raw-materials inventory, but this area paled in comparison to the excess finished goods inventory at the factory and branch locations. "We used to forecast by spreadsheet and by product groups such as passenger radials," Harris explains. "The passenger radial group was made up of probably 130 different sizes, but we used to just forecast for the group as a whole." The company's marketing staff would work the numbers, project what they would sell - for example, 85,000 passenger radial tires - and then manufacture a cross-section of sizes.

On top of the rudimentary forecasting process, other quirks in the manufacturing and marketing silos caused inventories to steadily balloon. "We had a group of employees working on the factory floor who were paid, and thereby motivated, by the number of units they could manufacture," Harris says. "We have certain products that manufacture far easier than others as well as products that produce markedly different yields from raw materials." For example, two different tires may be similar in size and appearance, but one tire may produce a yield of 96 percent per raw materials while the other produces an 82 percent yield. "The why of it is not always clear - all you know is that when you make this particular style of tire, you don't get as many tires through the hole in the wall where they keep score." Accordingly, on those days when they had a choice of which kind of tire to make, the workers made the one that gave the better yield.

On the surface, this looked great. "All charts management put on the wall said that waste was down, yield was up, production was up, the number of tires through the hole was up, and those were our motivators," says Harris.

Add to that equation the commission arrangement for the sales and marketing team, and it's easy to see how the company got into trouble. "The salespeople out in the field have a wide range of tires to sell. Some tire styles everybody wants, and therefore the commission the salespeople earn on those tires was very low," says Harris. Conversely, all those other tires for which there was low demand paid high commissions, which prompted the sales force to go out looking for markets in need of the higher commission product lines.

Inventory Pileup
"The problem developed when a salesperson would be working on a prospective deal for the higher commission tires and come back to the plant with 'I have a guy on the verge of selling, and he needs 60 of these earth-moving tires, so make them as soon as you can,'" he recalls. Then the deal would fall through, and Dunlop was stuck with the inventory. "We have tires lying on the ground in Johannesburg that have been there for six years. Our man promised us that they were sold, but when he submitted them he lost the deal for one reason or another. And then the tires were only good for throwing away because, for example, they were 141/2" rims and not used anywhere else in the world. And we made them on the motivation of a person who was driven by the commission he hoped to earn."

Occasionally, both production and sales/marketing combined to exacerbate the problem. "Manufacturing is told to make 200 of a particular unit, but they run off 600 because they get a better yield if they keep the line running. Everything in the factory looks good - they are above plan, everybody's popping champagne corks and enjoying their success. But when your motivations for manufacturing a tire are seldom directly related to what the market wants, you end up with a huge inventory of stuff, much of it unusable or in very low demand."

Coordination problems also contributed to missed external opportunities. When the market wanted tires the company didn't have in stock, the factory worked those tires into the production mix, by which time the market had changed. "A buyer in Scandinavia places an order for four containers of snow tires, and we deliver them smack bang in the middle of spring, at which time they are useless to the buyer," says Harris. "We were just simply unable to make the right product at the right time because the factory was busy on its own mission."

The matter came to a head in 1999 after the factory production and marketing teams stopped talking to each other because neither was ever satisfied with what the other was doing, he explains. "A few of us put our heads together and said surely, with the factory space and equipment we have, we can make the right tire," Harris recalls. However, the project design and implementation had to be done by a new Logistics Division, which was created specifically to stand between marketing and the factory.

Dunlop learned that change doesn't come easily to established companies accustomed to an Old World environment. "We weren't a new company that had the ability to quickly capitalize on opportunities and to apply new technology immediately," he says. "We came with a lot of baggage. And we had never addressed this issue of inventory control and forecasting - real forecasting - so we were looking at a completely new and very dynamic environment. We were at a big disadvantage in that respect."

Many companies have had at least some experience with this kind of information technology and therefore had a basis upon which to make comparisons. "But we had nothing to compare it to. For us, everything in this arena was a revelation. We were only one step away from doing our forecasting on a piece of paper with a ruler and a calculator and a pencil. We did have it on a spreadsheet, but in effect we were still using purely a flat piece of paper to do our forecasting."

Dunlop did not issue a formal request for proposal in its search for a technology partner. "Obviously South Africa is a much smaller market than the U.S., and the technology we were after, while it was implemented in a number of places, was not commonplace here," Harris says. After some initial discussions with software vendors as well as the company's mainframe supplier, Harris and others within Dunlop had a pretty good idea of which vendors to investigate further. Working together throughout the process, Harris and Dunlop's director of finance evaluated six companies, which were narrowed down to two finalists that were asked to make full presentations.

"They took our data and made their proposals, telling us 'this is what you've done with manufacturing and inventory, this is the history of your sales, and if you had done it properly, you could have gotten your inventory down to here and your sales up to there," says Harris. "The numbers were horrible. And this is where I believe a lot of these two vendors' salespeople lose their sales, because they go out there and tell their prospective customers the truth. They'd probably do a lot better if they told the companies that they were only half as bad as they actually are, and the shock would be more manageable."

However, once Dunlop faced the music and everybody's blood pressure returned to normal, the decision between the two was easy. "One of the two vendors got overly greedy and gave us an extremely high price tag, as they could see that we could save a fortune with a decent forecasting system," says Harris. "When we asked them to break that down in terms of costs, they declined."

It was different with Ability. "We were particularly impressed with their professionalism and the level of understanding they had about our business," says Harris. "Within a day of being on site, their people were talking about our product line in detail. Whether this is a discipline they have or because they predominantly used engineers as their software technicians, they had people who could walk into our factory and recognize machines that run off steam and machines that run on electricity, could differentiate an in-line process from one that was not in-line, and could identify batch processing from a continuous process. And it's important to have those skills, because those people needed to take what we had on the floor and apply it to their software."

The cost of the project was substantial, but when presented with details and projections, there was no hesitation on the part of the company's board of directors, Harris adds. "Everybody knew that we needed it."

A contract was signed Sept. 1, 1999, a server was ordered, and the Ability software was operational before the end of the month. Implementation for the finished goods sector of the company took approximately five-and-a-half months, though another 45 days was spent on tangential issues.

"We went from using a projection of gross sales to forecasting on a per-size and per-style basis, with Ability telling us what we ought to be making in every stock item we produce," he says.

The effect of making the right tires was extremely dramatic on inventory. "Initially our inventory started going up because suddenly we were making tires that we didn't have in stock and we couldn't sell the tires we already had in stock because we had made the wrong tires. But Ability said 'trust us and do it,' and sure enough, the inventory levels soon began to plummet because we stopped making tires we couldn't sell."

Instead of making additional quantities of slow-moving items, Dunlop moved excess inventories of those items between branches as necessary to meet prevailing needs. "We decimated our stockholding, and that generated cash," Harris points out. "Folding cash just came pouring in, day after day, week after week, for the first six or seven months. And we're talking about tens of millions of dollars."

Needless to say, Dunlop found this highly encouraging and quickly moved to extend Ability's Inventory Optimization to the company's raw materials supply as well as the maintenance and repair operations (MRO). Those inventories consisted of nearly 18,000 stock items such as machines, bearings, belts and similar items. Since Dunlop had gone through this process with finished goods, it made the implementation a lot smoother on the far more complex MRO side of the equation.

Nonetheless, implementation came as a serious shock to those accustomed to Dunlop's traditional mode of operation. "We started this project in MRO and all of a sudden all these people came running out from the ground floor - guys you've never seen before - asking why we were interfering with their jobs of buying materials," says Harris. "It clearly ruffled some feathers, because you suddenly find people with purchasing responsibilities who have been working based on rules of thumb that often had no scientific basis whatsoever."

For example, Dunlop traditionally would maintain an eight-week supply of an item because the lead time for that particular item was six weeks, which appeared to give the company two weeks of safety stock - at least that was the logic embraced by the particular purchaser. "We then asked how regular the supply shipments were, and they'd answer 'like clockwork - we order this Tuesday, and we get delivery on Tuesday in six weeks time. They simply didn't realize that they didn't need eight weeks worth of stock, only two."

Money Tied Up
All in all, it was an exciting time, he recalls. "It turned into quite a melee, with dust and fur and fists and people's pride and egos coming into play. Certainly in the raw materials operations, the people in charge of the store had no recognition whatsoever that they were holding too much stock. All they were concerned about was never running out, and we were trying to make them see that there was a lot of money lying about in the storeroom that could otherwise be used to buy more machines and make more tires."

Overcoming the human reluctance initially was quite a challenge. "A particular item may be mission-critical, but if we can buy it over the road, we don't need to keep two dozen on hand. And we had situations where precautionary measures went overboard." For example, the factory uses a certain type of knife to trim rubber from large pieces being prepared for the manufacturing process. "We would have 30 knives in stock, they'd use up two on the factory floor, and the guy in charge of inventory would order another dozen. They'd use four more the next month, and he would order another two dozen, because usage doubled from the one month to another," Harris recalls. "The stock of knives just never stopped going up." On top of the misplaced logic, there were way too many people involved in the MRO space, all driven by the same focus of ensuring that the factory never ran out of the items for which they were individually responsible. "There was virtually no thought given to the amount of money tied up in inventory."

Working with Ability, Dunlop determined a logic for replenishing both fast- and slow-moving items. "We attached criticality values to each item on the inventory list, determined product and material lead times, and fed the information into Ability. The software then told us what we needed to keep on hand." And the savings continue to pour in.

As Dunlop's familiarity with Ability grew, the tire manufacturer began publishing Ability-driven reports on the company's intranet so that replenishment and shipping activity was available for all authorized persons. "If the people at our Port Elizabeth branch want to know what they can expect on the next truck or what we are manufacturing on a particular day, they now can access that information on our intranet," says Harris. "If someone wants to know what imports of tubes from Taiwan are now on the water, and they want those numbers separated by 13" and 14" tubes, that information is available." On the maintenance side of the equation, "a chap can go into the system, target a group of machines, hit a button, select a specific machine and see every single spare part available on site for that machine plus any that are on order or should have been ordered."

In fact, he says, with the additional exposure via the company intranet, the software has become personified within the Dunlop organization. "People within Dunlop talk about Ability as if it's a person who has worked here for a long time, as in 'we can ask Ability what we should be keeping,' or 'Ability will tell us what to do.'"

There's a lot of interaction between tire manufacturers in South Africa, he says. "People talk at parties, at meetings, at launches, and we listen. Time and time again we hear from our competitors 'you can't believe what our forecasting is like.' Some guy who used to work in the factory and lost a couple of fingers in a machine becomes a forecaster because the company wants to keep him out of the factory. The level of professionalism in forecasting in the tire industry in South Africa is abysmal, and that, for us, is a great thing, because we believe our system truly is on the cutting edge."

All in all, the software has had a dramatic effect on the company's performance, he adds, far beyond their expectations. "The numbers are huge - they look like telephone numbers to us. And the performance benefits we have achieved are huge," says Harris. "Our ability to put the opposition to the post simply because we more often than not have the right tire on the right dealer shelf at the right time gives us an edge over the balance of the market."

Between March 2000 and December of that year, Dunlop gained market share in South Africa of about 4 percent, he adds. "We traditionally had about 29 percent of the local market. We're sitting now with nearly 34 percent, and the only reason to which we can attribute that gain is formalizing our forecasting process. And that represents to us a massive benefit.

"Our customer is the dealer, but effectively the person we want to reach is the consumer. In order to get to a wheel position, which is our ultimate goal, we have to have shelf space at the dealer. With Ability, we now get that shelf space, because we now provide the tires the dealer can sell."