Retailers are increasingly besieged with claims about the benefits of new technologies and practices for buying, allocating, planning, replenishment, promotional planning, pricing and other key merchandising activities. This article by Karabus Management scrapes away the hype about merchandise optimisation and looks at how retailers are boosting margins and inventory productivity by better predicting and meeting demand.

How often does a customer enter one of your stores and leave again without having made a purchase? Why didn't they purchase? Couldn't find the right item? The right size? The right price? How much does it cost your company to let potential customers walk out empty handed?

Now take the flip side. How often does a customer buy something at a sale price when they would have paid full price? Perhaps it was a company-wide markdown on items that were doing well in specific geographies but not in others. How much does that cost the company? And why does this happen?

Because from a headquarters location, your merchants cannot see traffic patterns and inventory positions in every single store. And how often do you have to "sell" residual product at the end of the season to pennies on the dollar because it did not sell at any price?

Now imagine putting a buyer, planner and pricer in your stores - every one of them - even if there are hundreds or thousands of locations. Every day, perhaps every hour, these analysts would be on top of every category and every SKU, making rapid-fire decisions to optimise each store.

The result? Better pricing. Less unsold inventory. Higher inventory turns. Fatter margins. Customers, conditioned to getting what they want when they visit the store, would become more loyal.

In essence, this is what a number of retailers are beginning to do - without of course, the impracticality of adding skilled analysts to every store. By implementing optimisation technology, supported by changes to processes, culture, roles and accountabilities, merchants are getting much more accurate, timely and granular information about consumer demand.

They are then using those insights to make much better decisions across their entire merchandising operations: what merchandise to purchase, how to plan and allocate it, how to price it (both initially and for markdown), how to promote it and how to replenish it on the shelves.

It's called Merchandise Optimisation. And it's all about making significant changes in the way merchants plan, promote, purchase, allocate, replenish and price products - changes that fly in the face of tradition but which are essential to generating substantial business benefits.

Without making those changes to process, accountabilities and philosophy, the goldmine of information the new technology delivers will remain untapped.

Retailers that are using the technology - and are changing age-old merchandising practices, further defining roles and skills, and creating accountabilities - are getting impressive results.

Northern Group Retail experienced substantial payback during the 2002
Holiday season, just one month after implementing markdown optimisation technology, supported by revised roles and processes. JC Penney has made major undisclosed improvements in gross margins after instituting merchandise optimisation.

But these are just the early returns. The technology is relatively new and evolving quickly. The changes in merchandising practices, roles and measures necessary to capitalise on the technology are just starting to be understood.

But while it is key to use the technology to provide crucial insights on customer demand, retailers must use this information to change business processes and get buy-in, in order to deliver substantial improvement in margins, inventory turns, comp store sales and thus earnings.

The point is not that one should be swept up in the euphoria of merchandise optimisation. Rather, it is to understand how to bring this opportunity to life and make it real for retailers.

Understanding industry best practices and how they can apply to each retailer's specific business is the key place to start. In certain cases, we have found that retailers can make substantial financial improvements in merchandising processes just by tapping the hitherto unknown functionality of their existing information systems and applying some best practices.

An example of this was a large apparel retailer who hired our firm to help it figure out how to deal with the 7 out of every 10 people who had gone into the store with the intent to buy and had not purchased something for the reason that "they could not find the right size" (sound familiar?).

Within 45 days, we identified some size optimisation features in its existing allocation software that it was not aware of. When this was implemented, together with merchant training and related process modification, it resulted in a more store-specific allocation, thereby significantly reducing this problem.

But whether or not new or existing technology is required, retailers must begin with a business - not technology-driven - approach that focuses on the greatest "pain point" in merchandising and what the potential returns will be.

What's driving retailers toward merchandise optimisation?
Three forces are driving retailers to get far better at the complex challenge of matching customer demand with the supply of the right items at the right times in the right places:

  • Lack of economic growth in retailing
    The economic downturn in recent years has forced retailers to fight over a pie which is shrinking or, at best, not growing (retail market share has become a "zero sum game").

    Even the world's strongest retailers are having trouble gaining a bigger slice of the action. In 2001, the average annual growth rate of the top 100 retailers was less than half the rate of the previous five years (4 per cent vs 9 per cent).

    Growth in certain retail segments has been flat or declining. In fact, sectors like specialty apparel have been shrinking. For example, between 1992 and 2001, US women's clothing stores' share of US retail sales (excluding restaurants and automobiles) declined from 2.6 per cent to 1.7 per cent.

    All this means that retailers must steal share from one another to grow which, in turn, means better assessing what consumers want, when, where and at what price.

  • The rise of the "process killers"
    Remember when the "category killers" were regarded as the biggest threat to retailing? Judging by Wal-Mart's success in toys (it now sells more toys than former industry leader Toys R Us), jewellery and photo finishing, to list just a few categories, we now know it isn't necessarily true.

    But a different retailing juggernaut has emerged over the last decade. We call them the "process killers." These retailers run circles around competitors because no one outperforms them in executing a key retail business process.

    Wal-Mart is the master of inventory management. Why has the $200 billion retailer raced ahead of Kmart over the last 20 years? It's not because Wal-Mart sells radically different products than Kmart. It's that Wal-Mart is much better at managing those products - at buying, allocating, replenishing, and pricing them.

    Few apparel retailers beat Spain's Zara for identifying new fashion trends and getting product in the store faster.

    Following the lead of such companies, other retailers have begun to recognise the importance of continually raising their process performance, particularly in those processes that directly touch the merchandise. And they realise that the "performance bar" keeps rising.

  • Rising consumer expectations
    Consumer expectations change faster and vary more from region to region. The shelf life of retail products like fashion apparel, toys, seasonal products and consumer electronics has fallen precipitously. Selling seasons at some apparel chains have increased from two to seven or more a year.

    In addition, what customers demand in Des Moines may have little resemblance to what they desire in Denver. All of this works against the DNA of the chain business model that has dominated retailing over the last century by bringing a common approach to store operations, advertising, purchasing and supply chains.

    Today there are significant disparities in customer demand for stores in the same state, even in the same town - and those expectations change faster than ever. As a result, matching demand with supply at hundreds of stores across a country is now exponentially more complex. Where it once was appropriate to run the chain one way, now it's store by store. And where once it was acceptable to revisit key decisions every month, now it's daily or at least weekly.

The revolution in demand forecasting technology
What's the key to achieving the promises of merchandise optimisation? Forecasting.
Getting the right product in the right place at the right time and at the right price requires an accurate understanding of the interactions between supply and demand as mitigated by price, availability, presentation, promotion, and time.

The good news is that forecasting has gotten better. Much better. Whereas the only forecasting available to retailers in the past has been "re-tooled" supply-chain approaches which were force-fit to retail, now there are very specific forecasting alternatives that operate down to the store level.

Key in these advances are the scalability and flexibility of forecasting approaches, not only relying on pre-selling data, but updating dynamically based on actual sales.
Fashion forecasting, which was once considered an oxymoron, is now possible, and it is driving the leading assortment, allocation, and markdown optimisation applications.

The five ROI drivers
The rewards of Merchandise Optimisation can be immense. On the other hand, there are plenty of examples of multimillion-dollar merchandising systems investments that don't pay off. Poorly designed and unrealistic project plans, resistance from merchants and store personnel, and ill-conceived pilots with mediocre results can all lead to an unfortunate justification for not proceeding. Five key drivers that will meet or exceed the desired ROI from Merchandise Optimisation:

  • Merchandising Process:
    Making fundamental changes to processes used by merchants, planners and others in the merchandising group by adopting key best practices most applicable to each retailer's business;
  • Store-Level Decisions:
    Basing decisions on store/SKU -specific data in real time rather than at chain level, on averages or gut-feel;
  • Bite-size Chunks:
    Breaking the initiative into "bite size chunks" that can be implemented quickly and with relatively quick results, thus also helping to fund other important initiatives;
  • Focus:
    Redefining roles, necessary skills, metrics and performance management plans within the merchandising group to create focus and accountability. (Moving the emphasis from "comp store sales" to "comp store gross margins" requires aligning performance measures and rewards with that new direction); and
  • Stores:
    Preparing store personnel for the operational impact from changes to markdown and promotions policies, signage and visual presentation.

Data from past fashion seasons is being productively employed to make smarter decisions about upcoming ones. Data from the current season is being deployed in near real time to make decisions about seasonal transitions. Forecasting has come a long way.

No one size fits all
There isn't a single approach to merchandise optimisation for all retailers. The right path depends on: a) whether a retailer is selling short or long life-cycle merchandise (see exhibit 1), and b) on each retailer's particular "pain point."

Merchandise optimization at short lifecycle retailers
Retailers selling short-lifecycle products (such as fashion apparel, toys, seasonal product and electronics) should focus on addressing such activities as assortment planning, store planning and allocation, promotions planning, supply chain and markdown management.

Markdown management is the first area of attack for many retailers. And that's for a good reason: markdown optimisation can generate improvements in margins significantly and rapidly and can also fund longer term initiatives. It has become a necessary tool, especially in apparel and other short life-cycle retailing, where "fresh" product has become crucial and getting rid of "stale" product requires more frequent clearance sales.

Apparel selling seasons for some retailers are no longer measured in months but in weeks and days. Markdown optimisation relies on advanced fashion forecasting techniques that determine the "natural demand," or the inherent pattern of demand for an item independent of seasonality, price, promotion, and inventory position.

Once the "natural demand" curve has been established, managers can stage "what-if" scenarios. "If I want to sell out of this line in three weeks rather than six, what price do I have to sell it for this week, and what will my total gross margin be at the end of the season?"

This same underlying analysis can also be used to evaluate promotional pricing strategies and take into account item-level, store level, and chain-wide promotions as a driver of demand.

Merchants no longer need to make pricing and promotion decisions based on instincts. They can get a much better handle on the impact of their pricing decisions on many more items because the technology is doing the "heavy lifting analytics" that it would take hundreds of pricing analysts to do using spreadsheets and calculators.

With more accurate, timely and specific forecasts of consumer demand, retailers can also improve the next link in the merchandising chain: store allocation. Here again, allocation personnel increasingly need new technology to manage the enormous complexity of ensuring that thousands of products are shipped at the right time in the right amounts to hundreds of stores.

Lacking technology to handle so much information, allocators have been forced to depend on crude averages which often lead to sub-optimal allocation decisions.

For retailers ready to attack the problem at its root, there is tremendous leverage in the initial plan. Few retailers are currently planning at the store/SKU-specific level, yet this is where the demand actually is.

Advanced forecasting and optimisation technology is also being leveraged to inform the initial plan, determine optimal breadth and depth of merchandise, optimal case-pack configurations, optimal flow patterns, optimal size profiles, etc - all at the individual store level within operational constraints.

Leading the charge among the short lifecycle retailers is JCPenney, which is applying forecasting and optimisation technology to markdowns, allocations, and assortment, and whose recent performance has demonstrated a significant return on its focus on revamping its merchandising processes.

Merchandise optimisation at long lifecycle retailers
Retailers of long-lifecycle goods such as basic apparel can benefit from merchandise optimisation as well. Typically, the areas of replenishment, initial pricing, assortment, promotions, and store space utilisation are the key opportunity areas.

While long-lifecycle retailers don't have to focus as much on consumer fads, they do have trouble staying in-stock on predictable items. Out of stocks have been estimated to cost retailers 3-4 per cent of sales every year.

For these retailers, merchandise optimisation is also about establishing a better initial price by more deeply understanding its initial elasticity. By this, we mean how price changes increase or dampen consumer demand, and what price optimises profits. Technology now exists to get a more robust view of initial price elasticity.

Retailers of long-lifecycle items can also improve replenishment and optimise store space with new predictive technologies. Wal-Mart is legendary here. Its store and supply chain systems enable it to review customer purchases immediately and react with lightning speed to sales trends.

For example, the day after the September 11 terrorist attacks, Target saw a spike in sales of American flags and tried to buy more of them. Kmart took three days to recognise the pattern and then tried to do the same. Both were stymied. Wal-Mart had beaten them to the punch. The very evening of the attacks, Wal-Mart saw the pattern and quickly cleaned out the entire US inventory of American flags.

Wal-Mart is making daily improvements to its forecasts on basic stock replenishment to generate better buying, allocation and pricing decisions at each store. Such changes not only improve margins, they free up buyers to look for exciting new products.

Making it real
Merchandise optimisation can have a profound impact on a retailer's profits and revenues. But getting those benefits is not simply a matter of implementing new software and training people to use it. Significant challenges stand in the way, all of which can be overcome.

While the requisite changes of merchandise optimisation will vary in type and extent in each activity, collectively they are sweeping. For example, in markdown pricing, merchants may have to change the look of price tags and how the product is presented in the stores; create or make significant changes to some roles; agree on "business rules" to optimise the computer generated recommendations; and integrate the new practices into the way buyers plan and take markdowns. Often, changes to performance management are needed.

Expert Analysis

Merchandising optimization: the right product in the right place at the right time.
Merchandise optimization helps apparel companies use demand information at the store and SKU level as a basis for strategic merchandising decisions in order to drive revenues. What and how much should I purchase? How do I allocate it? What price should it be? How should I plan for markdowns? What promotions should I run? When and how should I replenish the products? This management briefing, exclusive to members of, offers a great insight in to assortment, allocation, pricing, promotional and markdown: the key areas to boost sales, profits and price image. Click here to download this report from the Members Club.


In promotional planning, the changes to business practices are far more substantial given that this activity involves distribution centres, marketing, stores, and merchandising. Parties in all those areas must agree to play by the same set of rules, operate in a new process, and manage a different way of choosing and presenting products.

Buy-in, recognition of corporate culture and how to design the initiative to support and enhance the brand are all crucial to success.

Another key challenge is sizing a merchandise optimisation program correctly. The operational changes extend across a range of functions and departments. Retailers must scope out those changes in advance but implement in them "bite-sized" pieces.

Fashion retailer Northern Group has a massive merchandise optimisation overhaul underway. But it chose first to focus on markdown pricing before embarking on assortment planning, allocation and other fronts because markdowns accounted for a significant per cent of its gross sales.

A more formidable challenge in many retailers has nothing to do with breaking down the project into manageable pieces. It's about getting merchandising personnel to rely on computers rather than gut. In a retailing world of growing complexity, using intuition and last year's sales history to make decisions is no longer sufficient.

Overcoming the resistance of merchandising people requires taking the "mathematical" decisions out of their hands (pricing, assortment, replenishment etc.) and leaving them to focus on the "art" - selecting product, sourcing vendors, negotiating, creating promotions, and so on.

It also means assuring merchants that computer generated advice will be configured to use the business rules the retailer establishes. Ultimately, the decision resides with the merchant.

New ways of pricing, buying, allocating and other merchandising activities will create new roles and demands for new skills. Streamlining and automating manual work of buyers, planners, and analysts will mean focusing them on higher-value tasks, for which some may not be ready. A combination of training and/or new people will be necessary.

How those people are measured and rewarded may have to change as well.

Merchandise optimisation can be about increasing comparative sales, turns and/or gross margins. Retailers that reward their merchants on targets that conflict with where the biggest profit-generators are made - eg, continuing to pay for top-line sales growth when merchandise optimisation is focused on hiking gross margins - will face problems.

To make sure people are moving in the right direction, the organisation will have to institute rigorous measures of performance.

About the authors
The authors of this report are Antony Karabus and Stephen Granovsky who both work for US-based retail and consumer product consulting firm Karabus Management. Antony established Karabus Management in 1989, after eight years with an international consulting firm. Stephen is president of Karabus Management and leads the firm's Retail Operations and Merchandising Practice. Antony can be contacted on More information on the company can be found at: