The Art, Science, and Software behind (Optimal) Retail Pricing: Part 1

The “Four Ps” of marketing strategy, also known as the "marketing mix," are basically applicable to all businesses. TEC’s two-part blog post series in 2008 talked about the importance of pricing management in a down economy. Price and promotion in particular are the lubricants in retailing, although the two remaining Ps--product and place, are indisputably important there as well.

In his guest author article in Retail Info Systems (RIS) News, Wayne Usie, senior vice president of retail at JDA Software, remarks that one doesn't have to go far to see the impact the economy is having on retailers. The evening news is plagued with store closings, while "going out of business signs" and ominously empty “for rent” spaces seem to pop up on every corner.

Retail trade is one of the world’s most widespread activities. According to DemandTec's annual report, there are more than 1,500 retailers worldwide that have annual sales in excess of US $500 million. It is a tough time to be a retailer of any size, particularly today. For one, it would be an understatement to say that these are uncertain economic times, with the cost of living increasing nearly 5 percent per year and consumer spending growing much slower than the cost of living (inflation).

It’s Been Ugly out There

The ongoing recession has been raging for well over a year, and consumer anxiety continues to increase due to record unemployment rates (over 10 percent in the US), volatile fuel prices, the well-publicized housing crisis (i.e., US home sales have potentially hit rock-bottom), and the tight credit situation. Competition is also ever more extreme in the retail sector, with big box (discount or mega) stores proliferating, specialized niche retailers popping up, and ongoing consolidation everywhere.

This cutthroat competition means more options for consumers, which means that shopper loyalty is diminishing. Consumer behavior is thus changing toward increased price sensitivity and the advent of value shoppers. In other words, retailing is highly competitive and generally characterized by low profit margins. For more information on the current state of affairs in the sector, see TEC’s recent article “Retailers, Consumers, and the Recession: Weathering the Storm.”

Thus, retailers have made significant investments in IT, and most of these investments have focused on achieving cost reduction through increased operational efficiencies and transaction automation. These cost-cutting feats have been achieved by leveraging supply chain management (SCM), point of sale (POS), and marketing automation software, either painstakingly developed in-house or bought as packaged commercial applications.

AMR Research says that retailers are still spending their limited IT dollars in 2009. Retailers are still investing in IT despite the recession, but they are focusing on the following solutions and projects with quick and measurable paybacks:

  • merchandise assortment and space planning

  • regular price, promotion, and markdown optimization

  • in-store systems to improve the customer's experience

  • cross-channel merchandising

  • business intelligence (BI) that facilitates action

As a result of these IT investments, retailers have accumulated vast amounts of sales history data. While a number of academic techniques exist to analyze this data, incorporating advanced statistical analytics into a commercially useful solution that yields meaningful and actionable insights for retailers still presents significant scientific, engineering, processing, and cost challenges due to the vast amounts of data and the complexities of mathematical computing.

In plain English, sheer item volumes are overwhelming and manual pricing is tediously labor-intensive. Ineffective pricing strategies (e.g., based on corporate-wide blanket rules) and inadequate pricing IT tools (e.g., spreadsheets) additionally aggravate the situation.

Ancient Methods No Longer Work

Existing approaches that incorporate an understanding of consumer demand into retail pricing decisions generally have been limited to modeling sample data sets to provide limited insights. As a result, retailers have traditionally made merchandising decisions based on simpler--and often inadequate--approaches (examples below):

  • cost-plus or competitor-matching pricing (i.e., pricing that is exclusively linked to costs and/or based solely on competitors’ actions);

  • pricing that doesn’t consider demand implications

  • nation-wide pricing of items across the board, regardless of local consumer demand and competitive dynamics

  • “one-size-fits-all” assortments of goods, regardless of the unique preferences of consumers who shopped in each location

  • habitual (e.g., weekly or monthly “autopilot”) promotions, advertisements, mailers and other marketing programs

  • engaging business consultants to provide isolated category-based analyses

So how can retailers keep merchandise moving and revenue coming in? JDA touts optimizing product pricing as the key strategy, as the days of the aforementioned suboptimal pricing strategies that operated in a vacuum along with disparate systems or spreadsheets used for decision-making and execution are nearing their end. In its recent (and still ongoing) educational series of Web seminars, Revionics also pointed out that even if some pricing strategy exists, retailers are often unable to gauge whether it is working and when and why it was changed (if ever).

What Should Work, Then?

In today’s environment, retailers need scalable enterprise software that is able to model the numerous variables that affect consumer demand and process massive data sets in a cost-effective manner. Such a software solution must deliver actionable merchandising and marketing recommendations to achieve the retailer’s revenue, profitability, and sales volume objectives.

The leading on-demand retail pricing optimization and demand management software vendors, DemandTec and Revionics, define pricing optimization in the framework of a strategic pricing solution for retailers. Such a solution should consider the usual variables such as costs, competition, product mix (assortment), margins, etc., but also leverage POS demand (consumer demand) data to generate optimal prices.

Prices are then determined by proprietary math, economics, and statistics. The crux of the matter is to predict consumer reaction to price changes so as to generate optimized pricing for stores’ deployment, monitor effectiveness of pricing at the store level, learn from actual consumer behavior, and respond with even better price recommendations.

In other words, the strategic pricing solution should apply sophisticated mathematics, economics, and statistics to generate weekly base retail prices, as well as promotional and temporary price reduction (TPR) recommendations. Last but not least, in addition to being able to predict, generate, monitor, learn, and respond, the product must easily integrate with existing POS and corporate back-office systems.

To that end, JDA Software continues with its intriguing “edu-nouncements” or press releases (PRs) that try to educate the market and establish the software vendor as an expert to turn to. At the end of 2008, my two-part blog post series analyzed one particular JDA PR with concrete and credible advice to embattled manufacturers in a tough economy.

In July of 2009, JDA’s PR asserted that with their profits and revenue goals on the line, retailers are striving to build stronger bonds with their customers. One strategy that is reportedly gaining notice with leading retailers is consumer-centric pricing and promotions.

To that end, early-adopter retailers are pricing and promoting items according to local demand, consumer demographics, and other factors to maximize value. In order to survive in today's tough economic environment, retailers must tailor assortments specifically to consumer demand, as well as price and promote items according to subtle local preferences and competitive activity to realize maximum value. Successful retailers are connecting any price changes all the way back into their demand plans to ensure that optimum inventory is readily available for customers to purchase.

... But Modern Pricing Techniques Are Not Simple Algebra

As said earlier on, retailers compete for consumers who are becoming more knowledgeable, more selective, and, in many instances, more price-sensitive. Especially in the Internet and social networking era, consumers today devote considerable time to researching products and comparing prices prior to purchasing and have a greater array of choices in price, size, brand, color, and features. As mentioned earlier on, the growth of discount stores, warehouse clubs, and dollar stores, as well as the emergence of the Internet as a viable retail destination (e.g., the site that offers electronic manufactures coupons and free shipping for certain consumer goods) offer consumers further alternatives when purchasing goods.

Recent increases or fluctuations in the costs of fuel and other commodities have further increased the average consumer’s price sensitivity. For retailers to compete effectively, they need to better understand and respond to these changes in consumer demand and behavior through targeted pricing, marketing, and merchandising strategies.

Current industry trends such as new formats and competitors, cost, inflation (and other macro-economic factors), blurring of retail segments and channels, and changing consumer behavior all demand that retailers become more targeted and with more locally attuned offerings. More astute retailers will win in terms of same-store sales growth, market share, average basket size, profits, and lifetime customer value.

A basic principle of economics is that a change in the price of an item will affect the demand for that item. Every item in a store has a unique “price elasticity,” or sensitivity in terms of the change in sales volume based on a change in that item’s price. Small decreases or increases in the prices of some items may lead to significant changes in the demand for those items, whereas larger decreases or increases in the prices of other, invariably needed, items may have little effect on demand.

The first are high-elasticity products (or candidates for occasional price reductions), while the latter are low-elasticity products (or candidates for price increases). Indeed, pricing optimization is often invariably (and incorrectly) associated only with price reductions, whereas price increases can also be justified in many instances. Therefore, the gist of the matter, which is much easier said than done, is to use demand intelligence to scientifically measure consumer sensitivity to price changes.

No Item Sits Alone on the Shelf Either

To make matters even more complicated, changes in the prices of items in a store often have an impact on the sales volumes of other items in that store. This interdependence is referred to as the “cross elasticity” of demand.

Demand is influenced by a wide variety of additional factors, including store location, customer demographics, current store features (i.e., flyers and ads), discounts (e.g., buy one, get one free [BOGO]), in-store displays (e.g., aisle-end), the availability of complementary or substitute products, seasonality/holidays, competitive activity, supply/demand shock events (e.g., a natural disaster or a pandemic breakout), loyalty and marketing programs, and so on and so forth. These variables make calculating price elasticity for even a single item an extremely data-intensive and complex process. Calculating the cross-elasticity of demand for thousands of items is exponentially more difficult.

But the point is that no product can be priced in isolation. For example, a grocer selling a soft-drink product in a two-liter bottle and a 6-pack, 12-pack, and 24-pack assortment (some of which can be either in cans or bottles) needs to ensure that there is some distance in the prices between each item. Without a shared view of products, companies run the risk of under-pricing products.

Every retailer, whether it be an apparel, grocery, or consumer goods store, has key items that drive traffic into the store. Knowing what those items are and pricing them at or below the market value is critical to attracting customers into the store. Once in the store, there is a variety of convenience or complimentary products that customers might buy simply to avoid the hassle of going to another store. Those products can often carry a slightly higher price point because of the convenience of offering them at one location.

For example, a home improvement store might price power tools at a competitive price but may sell batteries and other accessories (e.g., tool bits) at a slightly higher profit margin to capitalize on the convenience of getting both items in one location (i.e., one-stop-shop). Revenue lost in one area can be recouped in others by capitalizing on cross-selling opportunities.

Price and cross-price elasticity are only two factors impacting an item’s demand. Accurate demand forecasts come from modeling many more causal (influencing) factors. In addition to those mentioned above, additional factors can be promotional events, markdowns, the item’s availability, cannibalization, affinity, new product introductions and discontinuations (e.g., the emergence of low-carb or pomegranate-based products), assortment, available space, etc. A model based on sophisticated mathematics and statistics can then be used to forecast future demand based on these factors, while individual drivers should be isolated for a post-event analysis.

Applying these economic concepts to make day-to-day pricing decisions presents enormous challenges to retailers of all sizes, particularly large retailers that sell tens of thousands of items and have hundreds, if not thousands, of stores. These retailers must determine how to price each particular item and whether to vary the price among different regions or individual locations. Retailers must determine the price of each item relative to competing products and the likely impact on their aggregate profitability if the prices of that item or competing items are increased or decreased.

Part 2 of this blog post series will analyze some common retailer practices, and explain the vernacular terms. Then the article will go into the building blocks of pricing optimization.

Your views, comments, opinions, etc., about any pricing solution mentioned above (and about the software category per se) are welcome in the meantime. We would also be interested in hearing about your experiences with this nascent software category (if you are an existing user) or your general interest in evaluating these solutions as prospective customers.
comments powered by Disqus