Revlon is a leading global manufacturer and marketer of cosmetics, skin care, fragrance and personal-care products. After several years of steady decreases in supply-chain inventory, the company hit a plateau. A breakthrough change was needed and in 2003 the company called on Deloitte Consulting LLP for help. Without any new investments in technology, the transformation team designed a business model that significantly reduced inventory in the first year and promises to repeat that performance in years to come.

Q: Tell us how you came to your present position at Revlon.
Laverty: I started at Revlon as a supervisor in a pick-and-pack distribution operation and spent nine years in the distribution, shipping and transportation piece of the business. Then I was given the opportunity to transfer over to manufacturing at our original plant in Edison, N.J. There I was in charge of several peripheral departments-promotional assembly, returned goods, warehousing and receiving. Also, we used to make our own displays in-house and that was part of my responsibility.

I next moved to Oxford, N.C., which was to become the site of our main manufacturing facility. I spent seven years there as senior director of manufacturing, running the manufacturing floor. One of those years was spent as the leader of a cross-functional team that oversaw installation of a new manufacturing system at Oxford and at our other manufacturing plant in Phoenix. After that, I went to Phoenix as vice president of manufacturing for two years before the company decided to consolidate all manufacturing into Oxford. I then spent a year as a project leader on that consolidation. My next move was to New York, where I am now working with the front-end of the supply chain, leading demand planning, materials management, and until recently, direct materials purchasing, which acquires all of the packaging and chemicals that we use to make our products. I've been in this position for the last three years and that's how I came to be involved in this transformation project.

Q: Can you briefly draw us a picture of your supply chain?
Laverty: Our major distribution center is in Oxford and is associated with our main plant. We have a satellite DC in Phoenix. We found that it is easier and more economical and efficient for us to ship full truckloads out there and then distribute from Phoenix to the West Coast states. We purchase a significant amount of direct materials a year and we have had various efficiency programs targeted at that piece of the business over the last couple of years. We were very successful at getting into strategic sourcing and really bringing the cost of our purchases down significantly. We had teamed with Deloitte Consulting in some of that work as well. So the logical place for us to go was the next largest bucket of cash, which was inventory.

Q: What was the situation at Revlon before you initiated this project?
Laverty: For several years in a row, we had been reducing our inventory at a rate of 6 percent to 7 percent a year. Then, over the last two years, our inventory as a percent of gross sales leveled out. We hit a wall, so we needed something to jump-start us and get that same yearly reduction rolling again. We felt we needed to make a major step-change, because working harder at the same things we were doing was not yielding the results that we needed. And we knew there were more opportunities. When you are managing a large number of SKUs with the associated significant value of average inventory, there are always opportunities.

So we thought we needed to really break the model we were in and build it back up from scratch. I think that one of the things that everyone is most proud of is that we approached this knowing that we wanted to make significant change without making a significant investment in systems. Even though the systems that we have in many ways are not optimal for our business, they do function and they do work for what we need to do. We did not believe we had to invest significant capital into new systems to get the answers we were looking for, so we wanted to make sure that our processes were sound and see what we could do just by attacking process.

Q: What are your primary systems?
Laverty: Our primary system is basically an Oracle platform onto which we attached some of our own legacy systems, such as our warehouse management. The MRP is Oracle and the cost system is Oracle-based.

Q: So how did you start attacking your processes?
Laverty: Everything starts with the forecast and the demand-planning group. You need to get the forecast as accurate as you can, because the plant will not build product unless it is in the forecast. So we started by taking a look at the demand-planning tool that we use, which is Manugistics. First we challenged whether we were getting everything we could get from that tool as it existed. With Deloitte Consulting's help, we challenged some of the algorithmic parameters, how we bucketed the information and how we delivered that information to the plant. We did some minor tweaks to the technology to have it better serve us, but it was nothing that we had to invest money in - it was just using the same tool in a different manner. For example, we split the forecast over a month evenly, 25 percent for each week, but we switched from a "roll forward" methodology to a "consume" methodology when we actualize the demand against this weekly split.

The next key was to use the information in shorter bursts. Previously, we would do a major re-plan once a month at the plant. So we would take a look at the forecast once a month and re-plan all our SKUs at that point. Then we would get exception reports based on takeaway of sales and react to those exception reports within the month. Now, we do it much differently. We have key SKUs-the top driving SKUs-that we plan on a weekly basis versus once a month. So we are looking at key SKUs in a much shorter continuum. We are not ordering materials as far out as before. And that has completely changed the way the planners at our facility do their job. The result is that we are bringing in less inventory.

Now, at the same time, we looked at the lot sizes we were running on the manufacturing floor and we made adjustments to that. Also, previously when we set safety stock levels in our MRP system, it wasn't done dynamically. It was static-you would have this much safety stock for an A item, this much for a B item and so on. This was based on sound MRP principles, but it did not use the system to dynamically adjust those safety stocks. With the help of Deloitte Consulting, we were able to apply some logic to our system so that now safety stock levels are dynamically changed based on individual SKU forecast accuracy. So now that happens on a systematic basis and in a way that is invisible to the planners.

Q: Did your planners like this change?
Laverty: We knew that this would be a major culture change because it would take our team away from doing something they had been doing for many years with a fair amount of success. Remember, this was a team of people that had brought inventory down consistently year after year. But in the most recent two years the improvement leveled off, so we knew we had to convince our people that we had to rethink what we were doing. So we took four very strong leaders that we had within our group and assigned them to the project full-time as change agents.

We knew that this was a big culture change and we needed key people-people who would be our future leaders-to take this business model and sell it, along with the changes in behavior that would be required. We also had strong sponsorship from our head of operations and we reported up to the CEO and CFO of the company. We kept them updated on a monthly basis, and that continues today, though now we are reporting more on favorable results rather than where we stand on the initiative. So we pulled these four leaders out of their normal day-to-day activity completely and then back-filled for them with some experienced executives from Phoenix.

These folks also had a team from Deloitte Consulting working with them. They gave us the ability to effectively crunch key data. If we had tried to do this on our own, we might have gotten there, but it would have taken a much longer time. So we needed the manpower and in-depth experience that they brought to the table. With their help, we were able to crunch data and develop business models and savings models a lot more quickly. This helped us get results sooner, but it also helped from a "selling the change" perspective because we could show the large amount of savings that were involved here. This was critical because we have close to 2,500 people in our plant and we were talking about really changing their behavior, particularly in terms of the lot sizes they were running. The manufacturing team naturally would want to run 100,000 pieces of black mascara all day long. And here we are saying, 'We'll save significant money in inventory by not doing this. We should expect to run 20,000 pieces and do it three times a month.' You would expect some push back, and being able to show what these changes would mean in terms of savings for the organization and in terms of converting inventory to cash was really important. You have to have a compelling story to get people to believe it is the right thing to do and it has proven to be very much that.

Q: Did you think of this as a "lean" initiative?
Laverty: We had a separate lean initiative happening at the same time on the manufacturing floor, so we just tied into that. Where we knew we would be going to shorter lot sizes, we used the efforts of the lean manufacturing group to focus on those areas so they could work at speeding up the changeover time. One fed into the other.

Q: How did you actually work with employees to get their buy-in?
Laverty: Our plant is broken down into cost centers. Running a mascara product is very different than running a lipstick, which is very different than running a liquid makeup. So when we did the business model and looked for the greatest potential inventory savings, we did it right along the same lines. Our business model and our savings model were based on the way the cost centers are aligned. To begin, we chose two pilot areas that we wanted to go after. These were makeup and lipstick. Makeup is a high-dollar product. Lipstick has a lower dollar value but we do a higher volume. Both are really critical areas. With makeup, we knew we could make a significant impact from an inventory reduction perspective. With lipstick, we do a lot of shades, so it is a difficult area. Another reason we chose these two product lines was because we were pretty sure we would have success based on the people who were planning and running those areas. And, in fact, the two pilot areas turned out to be very successful.

What we did was to meet with the manufacturing team and planning team and discuss with them what the goal was and the path we were proposing to get there, while asking for their input. We also laid out the potential savings and started to define their roles in the process. We explained how we thought it would change what they did on a weekly basis. Then we rolled out these two pilot areas and had a tremendous amount of success. We took the learnings from those areas and then started to roll it out across the manufacturing floor, across all of our multiple cost centers. Our game plan had been to do this over nine months, but the success was so compelling and happened so quickly that we did that in about six months.

Q: Do you operate more like a process manufacturer or a discrete manufacturer?
Laverty: We are a little of both. In our toiletry and hair color areas, we make bulk product upstairs and then feed directly from the bulk tanks into the filling lines. This is a high-speed operation. Our opportunity for savings on inventory was not quite as great there because we store in bulk tanks. We felt we had the greatest opportunity where we manufacture batches of products like mascara and makeup. We make the product upstairs in bulk manufacturing and then pump it off into barrels or containers. We plan the assembly line production and bring the bulk out on the floor when the packaging arrives. So this is a two- or three-layered manufacturing process, and we saw that we could realize the greatest savings by running smaller batches and bringing in our packaging in smaller increments.

This means we are not buying out as long. One of the problems that we have in our business is that the life cycles of our products are very short, so the further we buy out the more danger we are in of creating obsolete inventory because of swings in forecast. Colors and styles change quickly and winning in our industry is creating exciting, new products and delivering them to the market year after year. It turns over quickly, so you have to be very careful about buying long and investing in a lot of inventory.

Q: Did you change the way you measure performance?
Laverty: We always measured our forecast accuracy and we still do that using an industry standard known as the weighted mean absolute percent error, or WMAPE. The main thing we changed is that we went to a frozen manufacturing schedule one week out. So we are now measuring ourselves against what we decided on Wednesday of last week that we were going to run this week. This measurement is not just on total pieces, which is how we used to measure ourselves. Before, if we said we would make 200,000 pieces of lipstick and we made 200,000 pieces, then we did well. Today's measurement looks at whether we made the 200,000 pieces that we planned on making. So we are measuring how we did against our frozen schedule for the week. This is important because a lot of the costs and nervousness that we were experiencing in the system were driven by our chasing the out-of-stock list. Remember, I said we would plan once a month and then we would have exception reports that would guide our business. Now, we make a plan and execute against the plan and one week out that plan is frozen. We don't change that schedule unless it is absolutely necessary to do so and then the change has to be approved at a higher level. This takes a lot of that nervousness out of the system and forces us to execute against the plan. So we got away from the exception-report adjustments that were driving the manufacturing floor crazy with changes in schedules. A lot of people thought this would have a disastrous impact on our customer service levels, but that just hasn't happened. I am really taken aback by the progress that we have made in such a short period of time.

Q: What are your results?
Laverty: Well, as I said, we were stalled at improving inventory levels measured as a percent of gross sales. For the first seven months of this year, we have improved in this measurement by 20 percent vs. last year. Another way to look at inventory is in days of supply on hand. We have reduced days of supply by 26 percent compared to last year at this time. This is supporting the same level of sales, with a customer service level that matches last year's. So we have not given up anything from a customer service perspective. And we achieved these reductions without any major technological investments. Most importantly, I believe that we have established a platform to go much further because now this will feed upon itself. We are expanding the same principles into bulk planning and bulk manufacturing, so that is a new avenue. Deloitte has been disengaged for several months now and we are doing it on our own. We have tools and a framework that we jointly established that enable us to continue to successfully convert inventory into cash, which can be re-invested into growing our business. We couldn't be prouder of the group.