December 1997 Profit Navigation Models chart path for retailers to maximize profits through product mix management Vitaly Pyrih For years retailers have been adjusting their merchandise mix in attempts to improve or maintain profits: contracting or eliminating one category of merchandise from their offering; growing or emphasizing another. Some have shifted their product mix emphasis from hard lines to soft lines. Some have transformed their stores from broad-line department store merchandise to specialty stores. Others have evolved category dominant formats. And some have been engineered from the start to be "category killers." More recently retailers have been developing new mix strategies with the aid of data mining, market-basket analysis and other quantitative analysis techniques. They are addressing demographics and shopping behavior to create store-specific merchandising and targeted marketing strategies. Yet, the problem of maximizing returns through product mix remains largely unsolved because of the inherent difficulty of predicting how sales and expenses of a given category of merchandise will be affected as assortments are broadened or narrowed and as more or less space is allocated. What effects do specific increases in space, inventory and selling resources have on a particular business segment's sales, on selling costs and on expenses? How does its operating statement change? What happens to net income and returns at the total store level as investment and market share in selected categories are substantially grown or reduced? How do product mix changes impact net income and return on stockholders' equity? Observations A study done at J.C. Penney Co. Inc. prior to their repositioning in the late 1980s revealed that the company could potentially improve its net income by several hundred million dollars per year if they were willing to change their product mix. The statistical analyses were based on a sample of more than 500 stores, stratified by store sales productivity (annual sales per square foot of floor space) and grouped by similar expense structure due to market differences. Store profitability as measured by operating profit percent, store profit dollars per square foot and net income as a percent of sales were all found to be inversely proportional to the mean deviation of the individual segments profitabilities. Furthermore, the higher the sales productivity of the stratified sub-sample, the more pronounced was the relationship between mean deviation and store performance. The findings led to several interpretations and conclusions: Mean deviation is a measure of investment adaptation (or the lack of it). Stores that break away from "corporate" mix profiles and respond intelligently to prevailing market forces and trade area demographics by allocating more space and merchandise to their best performing categories, at the expense of the poorest performing categories, out-perform others in direct proportion to the magnitude of those changes. Mean deviation is a measure of performance improvement potential. By locating a store's profitability/mean deviation point on the correlation line, an estimate of the improvement potential can be obtained at various values of the mean deviation variable. Performance improvement potential is also proportional to sales productivity. Highly productive stores (having high sales per square foot of space) are the best candidates for remedial mix strategies. All else being equal, profit improvement potential is directly proportional to sales productivity and inversely proportional to segment mean deviation. Stores that have a wide mean deviation or "spread" in the rate-of-return of the individual business segments (e.g. department operating profit as a percent of sales ranged widely from very high positive to negative returns), generally have much lower total store performance. And, stores with the least spread have the highest profit performances. The concept of the relationship between total store performance and segment performance "spread" is illustrated below: Reallocating merchandise investment, space and selling resources between the two segments yields: To help better understand this phenomenon, an analogy between physical and economic relationships can be made. The profit improvement potential of a wide spread in segment profit performances could be thought of as being analogous to potential energy tied up in elevated water, or to the higher potential energy of a mountainous terrain compared to flat prairie. Higher peaks mean more potential energy. Larger performance spreads mean more profit improvement potential. If one wants to extract energy from this physical configuration one allows the mass at high elevation to be lowered to the valley floor. If one wants to generate additional profits one grows the most profitable business segments thereby reducing their individual rates of return. But these observations are not mere academic curiosities. The findings can be exploited by building models to act as windows; models to make visible the magnitude of the improvement potential, where it lies and how to reach it; models that help navigate the planning process to more profitable merchandising strategies. Models PC-based models have been developed that make for powerful merchandise investment planning tools. The models are augmented by algorithms that have the ability to calculate incremental changes in segment sales caused by incremental re-allocation of floor space, inventory and selling resources. The models also incorporate the effects of expense leveraging: they calculate changes in margins, operating expenses, profits, etc., caused by changes in segment sales productivity. The models are formatted like operating statements and balance sheets; one for each business segment. During a typical run, return on equity is calculated for each segment and they are ranked in descending order of ROE. Investment in the poorest performing segment is reduced by trial increments, and increased in the segment with the highest ROE. As this is done, the impact on sales, gross margins, operating expenses, net income, assets, equity and ROE is recalculated for the two segments whose resources are being adjusted. Again, all segments are ranked and the poorest performer gives up floor space and merchandise investment dollars to the best performing segment. Unless the contracting segment is "protected" from further contraction by a previously established management criterion or the "receiving" segment has had a ceiling imposed on it by management, each iteration simulates a re-allocation of investment. And for each iteration, all the operating statement variables are recalculated by segment and for the "store entire". The process continues in this manner until a specified program limit is reached, which controls the degree to which the process is driven. The model concludes with a recommended merchandise investment plan, a floor space allocation plan, and with sales and profit projections by department, and, store total. At this point, merchandisers review the plan and make further adjustments to the original set of contraction protection values and to the growth limits and the model is re-run. After several runs and with ample merchandising input, the sales and merchandise investment plan is finalized and passed to merchandise planning and allocation for implementation. Opportunity Adding a model to existing planning/allocation systems empowers retailers to make very big profit improvements. It is estimated that a "profit navigation" model used at the front end of a typical merchandise planning process/system can improve store profits by up to an additional 3-4 percent of sales. This statistic is based on numerous model simulations using actual data from many different stores and store chains. It is also supported by the findings of J.C. Penney Co. Inc.'s statistical study mentioned earlier and by the actual results in performance improvement that J.C. Penney Co. Inc. experienced after they changed their product mix. The following table shows the results of a model run for one store of a soft-line store chain in the southeast. Other chains, other locations, other types of stores — from drug stores to full line department stores — exhibit similar profit improvement potentials. The amount of improvement depends to a large degree on how far a retailer is willing to drive the process. J.C. Penney Co. Inc. might be considered a pioneer in the concept of "profit navigation" to improve returns by shifting the product mix assortment emphasis. By the mid-80s their performance had deteriorated: net income was just under $400 million and they had an ROE of about 12.5 percent. They changed their product mix, emphasizing the more profitable apparel lines and repositioned the company. Although other performance improvement actions were taken during this period, the next four years showed a remarkable improvement. Their net income grew steadily to over $800 million per year and their ROE rose to over 20 percent. The exact amount of profit improvement that can be achieved by a retailer implementing such an approach depends on the magnitude of the initiative and on the merchandise classification level that is being addressed. Store-specific merchandising and micro-marketing are ideal applications of the profit navigation concept. Other applications are in marketing to trade area demographics by reaching out to the targeted demographic sector with the most profit-maximizing assortments. These type of "what if" models could also be useful tools in the strategic planning and development areas of retail organizations when used for strategizing, brainstorming and visualizing future scenarios. The situation seems to be ripe for educating merchants in the use of quantitative techniques and for providing them with modeling tools. What Are The Obstacles ? The state-of-the-art of merchandise planning and investment allocation is evolving to where it can become a major business initiative in its own right. But are retailers ready for this paradigm shift? Are they ready to re-engineer the planning process, to traverse boundaries, to transcend turf issues and to challenge old organizational mores? Some obstacles are technical. It is likely that the notion of using incremental profits to plan category growth and contraction to maximize profits has come up many times over with retailers everywhere. Incremental profit analysis can easily identify the best performing categories of merchandise as candidates for growth and the poorest performers for contraction or elimination. Yet static incremental profit data can not be used in a dynamic environment. The crucial question is: What will the incremental become as the investment base is changed? Will the projected incremental be sufficiently high to warrant the investment change in question? It brings one to a fundamental question of sales elasticity: What are the expected incremental sales at all points of the incremental investment variable for a given category? Do incremental sales tend to keep up with the growth of investment or do they drop off rapidly? How rapidly? What are the conditions that influence these rates? To do a valid job of planning the product mix, projected or forecasted segment incrementals are required. This is an operations research problem that can be solved. And there are signs that retailers and consulting firms are now beginning to be address these type of issues with a deeper realization of their significance. But most obstacles are "change" related — veering from traditional ways of doing things. Financial performance alone of a category is not considered to be the ultimate test of its viability nor a necessary criterion for its growth, contraction or elimination. Merchants lay claim to a much broader decision-making dominion. They point to such concepts as loss leaders, i.e. certain categories although themselves not profitable, generate traffic which in turn contributes to additional sales in other categories. There are other rationales for keeping the status quo as well. Customer expectations are frequently cited as reasons for keeping certain unprofitable classes of merchandise. But if floor space in a store is a constant and few categories or departments are reduced or eliminated, for whatever reason, then little growth can be expected from those segments that would add more incremental dollars to the store. Traditionally, segment growth and contraction has been minimal. Rarely is segment investment proportional to segment performance. Profit Navigation - Using A Model The observations stated earlier about profitability and performance spread point to a solution to the mystery of maximizing returns through product mix. Retailers should be constantly on the lookout for individual stores or store groups that have a large variation or "spread" in performance between individual business entities. Wherever large variations are found, large profit opportunities exist. Highly productive (sales per square foot) stores with a large "spread" are the best candidates. At the heart of maximizing returns lie two simple principles. One is exploiting "diminished" incrementals by which an immense amount of extra profit can be squeezed out with investment planning and "segment" market share growth/contraction. It is an apparent contradiction to the law of diminishing returns which states that providing more floor space, more merchandise and more selling resources eventually delivers less and less additional sales. This law is not violated; it is just ignored for a finite range of segment growth which contributes to the total profitability of a store. The other principle is the concept of a driving force or an "engine" — the need to establish a process driver by enveloping the planning processes in a continuous process improvement methodology. The steps to higher performance are few and easy. First, construct a model and exploit opportunities by adjusting merchandise investment and business segment market share growth toward higher rate of return products or segments. Make maximizing total store returns the true objective, even if doing so degrades the performance of best departments/categories. The graph above illustrates the theoretical limit of the investment reallocation methodology. Some segment performances which can not be improved beyond a certain point (I, J & K), are eliminated. Performance of some segments is improved to a level of store "average" by weeding out poorest performing sub-classes within those segments. Other segments' performances are reduced to store "average" level through growth of offering. Note that the very process of improving total store rate of return will level the individual category performances. Those categories with originally high ROE's will experience lower ones and low ROE categories will either experience higher ROE's, or, be eliminated entirely. Second, encourage management to set up a pre-planning, guideline development function as part of a merchandise planning and allocation process or as part of a merchandising strategy development unit. Develop this area as a new discipline or new organizational unit; a unit that is responsible for model building, forecasting and merchandise strategy development technology. Lastly, get your organization to do it sooner rather than later, before the competition is also doing it. As the potential from automated replenishment, from re-engineering, from quick response, from vendor alliances, supply chain management and from other such endeavors is depleted, maximizing returns by "profit navigating" the product mix management process will become an ever more important area of profit improvement opportunity. And modeling will be the driving force behind the new wave of profit improvement initiatives. Vitaly Pyrih is currently vice president of inventory management at Brooks Brothers. Formerly with the J.C. Penney company, Pyrih has held various positions in inventory management and strategic development. He can be reached via e-mail at vitalypyrih@cs.com Table of Contents OR/MS Today Home Page OR/MS Today copyright © 2000 by the Institute for Operations Research and the Management Sciences. All rights reserved. Lionheart Publishing, Inc. 506 Roswell Street, Suite 220, Marietta, GA 30060, USA Phone: 770-431-0867 | Fax: 770-432-6969 E-mail: lpi@lionhrtpub.com URL: http://www.lionhrtpub.com Web Site © Copyright 1999, 2000 by Lionheart Publishing, Inc. All rights reserved.