Quantum Retail: Challenging the "Enterprise Apps Establishment" and Retailers' Mindset - Part 1

Every year-ending holiday season reminds us of the importance of consumer spending and the retail sector for the United States (US) and global economies. Many economists and pundits are then awaiting with trepidation the Black Friday sales outcome and reported consumer sentiment or Consumer Confidence Index (CCI) as bellwethers of the economy in the New Year.

While this past holiday season seems to have gone quite well for most retailers (especially in terms of their online business growth) according to a recent National Retail Federation (NRF) announcement, hardly any retailer can now relax and breathe a sigh of relief. Retailers are in a tough business and are constantly seeking tools to enable them to delicately balance their sales, inventory, and profit figures.

The current market dynamics demand that retailers do more with less while still zealously protecting their market share. They need to juggle a multiplicity of variables in terms of assortments that are finely tuned to local demographics, locations, and clusters of their stores and distribution centers, merchandize allocation and pricing, workforce levels, space planning, and so on and so forth in order to best align their business strategies with today’s ever more informed, empowered, and fickle consumer.

How quickly can retailers react to the hectic pace of change, and fulfill the local demands of their stores? How do they plan to meet the needs of new shopping channels (which represent more ways and opportunities to not only attract but also to disappoint the consumer), while meeting the challenge of internationalization? My 2009 series offered examples of some astute retailers that have managed to thrive even in possibly the harshest environment of late.

Merchandize Consisting of Ever-more Difficult Items

Have you ever thought of what kind of products are considered as difficult products for retailers to handle? Well, big ticket slow-moving items, sized merchandise, highly volatile selling items, seasonal products, short-lived products, perishable products, vendor pack-constrained merchandise, heavily promoted items, vendor-allocated (scarce) merchandise, and items with long lead-times all represent a bag of tricks for retailers. They indeed have a vast choice from which to choose their poison.

Consider that the majority of most retailers' assortments are made up of slow movers, which would be those stock-keeping units (SKU) that experience less than one unit of sales per store per week. Research shows that these items represent between 60 percent and 90 percent of most retail assortments.

Often, retail software systems are still executing ineffective Min/Max replenishment strategies for these items. Retailers often treat all slow movers similarly as far as placing inventory or perhaps they use arbitrary points where service level objectives change. To that end, what should a retailer do differently for an item that sells, say, 0.24 units per week, vs. one that sells 0.08 units? The question in these cases should rather be “How should I place my inventory?” The answer would be “With some strategy in mind.”

Namely, is the objective of offering this particular assortment to achieve a traditional 95 percent service level with two days worth of supply or to drive store traffic (i.e., maximize sales and minimize lost sales, although remaining subject to about 10 percent of waste)? Or, is the goal to make money (i.e., minimize waste and maximize profit), to offer a fringe assortment (to minimize waste), or to just fill stores’ available space with expected inventory? These different strategies can lead to different results for the same product with the same features. Even changes in metrics such as desired gross margin can change results of the same product with the same strategy.

This evaluation should happen constantly by item and by store with a deep understanding of the tradeoff between item availability (customer service level) and the costs that this held inventory creates for the business (this evaluation can also be represented as the trade-off between sales revenue and profit). An astute retail system must suggest a stocking level that best meets the strategy objectives, although the user can do manual overrides as necessary.

Yet many retailers are still falling into the trap of “It is all about the forecast accuracy!” since for slow-moving items the value of the forecast is significantly diminished if not irrelevant. The ability to forecast has lately become a question of psychology (i.e., “Will customer A come in on Day X for product Z?”) rather than sociology (i.e., the likelihood of a general customer to come in for that product).

Dealing With a Panoply of (Largely Static and Isolated) Systems

Therefore, retailers need new models and thinking as better answers for their merchandize assortments, item allocations, and store labor in order to handle tricky factors such as seasonality, lifecycle phases of their products, weather, promotions, commodity price changes, average day-of-week and time-of-day sale volumes, and so on and so forth. It is difficult to properly understand customer demand when information comes from isolated silos, i.e., retail-oriented point systems.

Merchandise and assortment planning systems typically recommend target figures for store capacity or assortment presentation quantities (although volume targets may also be used) based on point-of-sale (POS) historical data. Product lifecycle milestones (i.e., markdown dates) are more likely to come from some planning system (i.e., supply chain planning [SCP] or enterprise resource planning [ERP]) or retail-oriented product lifecycle management (PLM) system, but may be incorporated into existing replenishment systems. For their part, seasonality, size profiles, and store clusters data may come from a data warehouse (or from even more targeted point solutions), but may also exist in some Allocation or Replenishment systems.

Furthermore, feeds for advance shipping notices (ASN) and expected delivery notifications (EDN) come from a Trading Partner exchange or a supplier’s electronic data interchange (EDI) system and are often managed through a warehouse management system (WMS). Last but not least, many point products are used for setting everyday prices, promotional prices, and markdown prices (see more information my 2010 series).

Needless to say, sets of these traditional solutions for retailers rely on manual user direction to configure the system to solve problems. As an improvement, savvy optimization solutions can periodically review performance metrics and suggest system configuration based on historical or desired performance. Some vendors have been seeking and gathering ever more solutions that cater to retailers in an end-to-end manner.

Some examples of broad solution sets for retailers would be those coming from Oracle, SAP, JDA Software, IBM, SAS Institute, Epicor Software, RedPrairie, Retalix, Island Pacific, Jesta IS, Celerant, and so on. But these solutions have yet to be able to continuously learn from what they “see” and promptly adapt the system to reflect today’s needs, without the need for user direction or periodic review. Needless to say, these offerings also require very sophisticated users in order to produce promised results.

Enter Quantum Retail

This brings me to Quantum Retail Technology, whose staffers I first met at the 2010 NRF BIG Retail Show. Although not a well-established name in the retail sector due to its nascence and budding install base, the company’s staffers can talk about their vast collective experience. The company was started in 2004 by a group of retail software executives (many of whom are from former Retek, prior to Oracle’s acquisition) that were frustrated with the inability of existing technology to address the issues of the day and to deliver on competitors’ (over)promises of typical retail business cases.

Starting with a clean slate has allowed Quantum to develop new retail offerings without the constraints of existing technology and architecture and to apply new thinking and science to the problems that were outlined earlier. This has allowed Quantum to bust some of the aforementioned myths of retail software and business processes.

Starting with its Q:Allocation and Replenishment module, Quantum has since moved on to develop and implement Q:Forecasting and Order Planning and Q:Assortment and Range Planning. These product lines have been developed organically, without any acquisitions to date. The applications are developed in Java with a user interface (UI) that is accessed through a Web browser.

Quantum is based primarily in North America and the UK, with offices in both regions employing over 80 full-time resources. The product can be translated to any language, but is currently deployed only in English. Quantum currently has the following five customers:

  1. Guitar Center, a specialty hardlines (musical equipment) US retailer

  2. New Look, a fast fashion UK retailer

  3. Kohl’s, a well-known large US department store 

  4. Marks & Spencer (M&S), a UK grocery chain (clothing, furniture and homeware, flowers, gifts, wine, etc.)

  5. Matalan, another UK fashion retailer

  6. Eddie Bauer, a US outdoor retailer

There is a good, nearly even split between the company’s UK and the US customers, and a fair spread across the retail subverticals also. Quantum has sought to solve problems for all types of merchandise,including the most difficult inventory problems: the high-margin slow-movers and the very volatile sellers – i.e., the problems that are not being solved by traditional retail systems today. The vendor has worked with different partners or internal teams on each implementation, and there are no current long-standing partnerships.

Quantum’s Main Product Lines

Q:Allocation and Replenishment provides the answers to retailers in terms of how much they should be stocking, what items they should be asking for next, and when they should be asking for it. For its part, Q:Forecasting and Order Planning helps with regard to how much should a retailer buy and when.

Finally, Q:Assortment and Range Planning advises retailers on what they should be merchandizing and where they should be ranging it. All three modules sit on top of the Q Platform, which provides more than just a common demand forecast to drive the business. The Q Platform provides a common understanding of how each product is behaving in every store, why the product is selling that way, and how Quantum expects that product to behave in the future.

This common understanding is leveraged by every business process (i.e., forecasting, historical demand, behavior modeling, intelligent clustering, size optimization, pack optimization, markdown optimization, multi-echelon distribution, multi-objective inventory optimization [IO], assortment rationalization, and model selection) to drive smarter, more granular answers in an automated, exception-driven way. Critical here is the constant evaluation of multiple dimensions of merchandise and location. Quantum’s system identifies and uses the most relevant results, weighted to reflect significance, and to continuously update seasonality in addition to merchandize lifecycle and overall demand.

Part 2 of this series will continue with an in-depth discussion with one of Quantum’s executives. Until then, your comments and opinions with regards to typical retailers’ issues and solutions are more than welcome.  I would certainly be interested in your experiences with various retail software tools in general and with Quantum in particular.
comments powered by Disqus