by Jeff Sandgren
Some of the words we use, we hate ourselves for in the morning. ‘Optimization’ is one of them. ‘Optimization’ is to business-speak as ‘cute’ is to babies. It almost always fits in the conversation somewhere, but often doesn’t really tell you anything. It does serve to identify the skulking presence of Brand Marketers who expect you to believe that their solution is not only good for you, great for you, and better than any competitive offering for you – it’s optimized, and therefore the best for you that anything in its class could ever possibly be. At least until the next release.
But let’s not throw out that Cute baby with the Optimal bathwater. Optimization isn’t just marketing hype. It is a solid science whose time has come, as was evident at this year’s National Retail Federation (NRF) BIG Show where the perfect storm of empowered, value-conscious consumers and ravenous, stock-conscious investors narrowed the tightrope of retail performance to a razor’s edge. If academia’s motto is “publish or perish”, the New Retail’s motto may be “optimize or perish”.
“National” Retail Federation is arguably a misnomer anyway, as attested each year by the growing international presence. As long as we’re celebrating the 100th anniversary of this illustrious association, perhaps we should change the name to (dare I say it?) the Optimized Retail Federation. The NRF is dead; long live the ORF!
Semantics aside, pity the poor Retailer of today. How to satisfy C-level expectations for volume and margin gains over last year’s campaign when consumers have so much more knowledge at their fingertips? The arbitrage of shopper price ignorance is as much a thing of the past as the luxury of widespread shopper price insensitivity. To be clear, there still are, and apparently always will be, that segment of shoppers who don’t pay attention to prices, and just want what they want when they want it. But for them, the Merchant’s need to know is no less: now it is assortment, size, and inventory optimization that must be dealt with. No matter which segment you target with any strategy, that consumer has choices and knowledge like never before, and if you can’t optimize the driving characteristic that delivers the retail experience they crave and shape their shopping behavior, your competitor will be happy to do it for you.
Leveraging shopper insights has always been one of the core competencies of the most successful merchants and marketers; but in the past this was often more about art than science. At the NRF conference, DemandTec shared results of recent research fielded by RetailWire that polled nearly 600 industry respondents. Nearly 80 percent said shopper loyalty program data is the ‘most actionable’, and more than 90 percent consider the application of shopper insights in most business processes as a shared responsibility between trading partners – hence the current focus on collaborative solutions.
We set forth to review a solid sampling of optimization applications this year, and found plenty. Even more significantly, we found that these solutions are reaching a maturity and sophistication on par with their newly-heightened relevance. What follows is a list of highlights from conversations with some of the leaders in the field. Bottom line: if you’re a retailer, and you’re not utilizing these solutions yet, baby, your bathwater may be about to run out.
At the heart of any optimization approach is the key question: how do you determine the Best Solution to a given problem. Applied Predictive Technologies positions itself as a world leader in helping organizations harness the potential of “Test & Learn”, a powerful fact-based approach for choosing, targeting, and tailoring strategic and tactical actions.
The central premise is that modeling has to be done in the real world in small scale to validate. The key is a methodical creation of the control group, which should behave like the test group “when nothing is going on”, hence a “simulation versus null” test.
Generally, they evaluate multiple design approaches using simulations of each approach versus null test sets to determine each approach’s average error, and select the design with the lowest average error. If a design approach = a “well-accepted “established” hypothesis” then it’s assumed to be correct. For each test, multiple control ‘matches’ are determined, based on financial patterns and store attributes such as age, proximity, size, etc. If that sounds confusing, think of it simply as using past behavioral correlations to construct future test designs.
According to Scott Setrakian, Managing Director, 28 of the top 100 U.S. retailers, 10 of the top 25 U.S. restaurant chains, and 5 of the top 15 North American banks use their solutions. And the trend is upward, according to Scott, who told us “leading retailers increased their use of scientific testing by more than 21% in 2010 compared to 2009.” U.S. Retailer customers include Big Lots, Food Lion, Office Depot, Publix, and Staples. http://www.predictivetechnologies.com
DemandTec is a proven player in this space with implementations that connect more than 340 retail and manufacturer customers on the DemandTec network. This reflects their architecture and approach, with collaborative solutions built to enable shopper-centric merchandising and marketing solutions. DemandTec customers include Ahold USA, Best Buy, ConAgra Foods, Delhaize America, General Mills, H-E-B Grocery Co., The Home Depot, Hormel Foods, Monoprix, PETCO, Safeway, Sara Lee, Target, Walmart, and WH Smith.
Last year they launched DemandTec Shopper Insights™ solution, part of their overall ‘nextGEN’ platform. With these tools, merchants and marketers get even better information on sales trends by penetration, buy-rate, shopping trip statistics, and more, enabling them to target specific shopper segments with more tailored assortments, promotions and pricing. Within a month of the launch announcement, Target Corporation adopted it as an extension of their nextGEN implementation, and others are quickly following.
Complementing DemandTec’s analytics-based software services, the company recently announced DemandTec Connect™, which introduces powerful social messaging and collaboration capabilities. The company says these will “… make the DemandTec experience even richer and more collaborative for its community of customers and partners on the network.”
DemandTec is also one of the leaders who have expanded their solutions to serve the special needs of apparel and other short lifecycle products. Recent additions to this solution set include ‘rebuy optimization’, size profiling and pack optimization, and store cluster analysis.
To add even more horsepower to their platform, DemandTec just completed acquisition of M-Factor, whose predictive analytics software for marketing mix and trade investment spending is now part of DemandTec Decisions™. According to the company, their vision is “… to shift the industry toward a more dynamic, holistic, and collaborative planning model that drives better decisions, better results, and more value to retail trading partners and, ultimately, the consumer.” www.demandtec.com
We met at NRF with Heather Loisel, Senior VP Marketing. Heather is a recent addition to the senior management team. She recalled that, shortly after joining JDA, she attended their FOCUS Customer Conference, held each May, where she developed the strong impression that “JDA is the best kept secret in SAP implementations.”
JDA are repositioning their solutions from a historically supply-side focus to the demand-side focus that was a common theme in countless presentations and booths at NRF this year. They now speak a lot about the importance of “demand signals” and store-level assortment planning. “We are one of the few to offer end-to-end planning solutions,” says Heather. Their portfolio includes major acquisitions over the past several years of Arthur, Intactix, E3, Manugistics (the core competitive offering), and, last January, i2, strengthening their positioning as “The Supply Chain Company”.
Recent success stories include Ace Hardware, Hibbett Sports, Wilsons Leather, and Woolworths Holdings. Heather also cited recent success with Lowe’s and Whirlpool, in which JDA’s Collaborative Planning, Forecasting and Replenishment (CPFR) implementation “tied” Whirlpool supply to Lowe’s demand signal.
JDA differentiates itself especially on their ability and track record implementing solutions in the broad Enterprise Resource Planning space. “The world’s largest retailers use JDA,” says Heather, “but you don’t have to be large to be a good fit for our solutions.” www.jda.com
Another very established player in price and promotion optimization is KSS Retail. Their VP Marketing, Lyle Walker, tells us that “we started doing retail price optimization back in 1993, and there was already 15 years of prior science behind it then, so we’ve been perfecting this for over 30 years. KSS has been doing it longer than anyone.”
A major development for them occurred in 2010, when UK consulting powerhouse dunnhumby acquired them. Dunnhumby’s client relationships are exclusive within a given market space (geographic and class-of-trade), so it was necessary to make KSS an independent subsidiary to avoid conflicts.
They are “getting into” markdown pricing, but are not there yet. They claim to be implementing a unique and superior approach to the science of markdown, with near real time transaction-level data. Much focus now is on the Heartbeat® Shopper Insights Platform, a “very granular, intelligent sense-and-model solution” that determines when problems arise in store selling conditions. The pitch here is that the retailer can effectively use shopper transaction data with KSS’ analytics as a shelf auditing mechanism. Results are extracted into decision trees, which “take price model optimization to the next level”. The other key element of Shopper Insights is, as the name suggests, analytics across customer segmentations. Users can still see aggregated results across all segments, including cross-effects, but now they can drill down to specific targeted segments.
KSS website claims 32 major clients. New customers include O’Reilly Auto Parts, 7-Eleven, Fred’s Inc., and Foodstuffs (a New Zealand retailer – the dunnhumby relationship is providing more international reach as well.) www.kssretail.com
When it comes to retail price optimization, the common approach is to emulate the three basic pricing strategies of retail merchants: they begin with ‘everyday’ pricing, pique interest and demand at times with special ‘promotional’ pricing, and end the cycle (especially for apparel and electronics) with ‘markdown’ pricing. Taken as a whole, this is commonly referred to as Lifecycle Pricing.
According to Susan Boyme, VP Marketing, Lifecycle Pricing is Revionics’ core solution. They strongly position their integrated forecasting, an approach that “…uses a single forecast across all software modules to provide a more accurate, consistent version of consumer demand.” Historically, their solutions began with fast-moving consumer goods; but they have branched out, asserting that “our science adapts to other merchandise categories with different characteristics.”
Their approach is to take retailer (and syndicated) data, and make recommendations on pricing. This is a key differentiator for them. Their solutions are about finding and recommending what to do, and ranking the items and actions which will help meet the retailer’s objectives.
Revionics is one of the larger, more established players in price optimization, claiming over 60 leading retailers as customers, with over 20,000 locations. Recent customer additions include Dick’s Sporting Goods, Roundy’s and the Tractor Supply Company. www.revionics.com
Diana McHenry, Director of Global Retail Product Marketing, shared with us SAS’ excitement over two ‘big news’ announcements. The first is their positioning of their new offering as next-generation “high-performance predictive analytics”, with their “turbo-charged SAS® in-memory computing platform”.
First go-live customer is Macy’s, as announced by EVP Steve Nevill in a Big Ideas session, scheduled to happen later this year. Another undisclosed client is positioned soon thereafter. Larry Lewark, CIO Macy’s, says “we view this as a major breakthrough for quick analysis”. They are already claiming results for improved performance, including “10 times reduction in demand modeling and forecast cycle time, a 3 times improvement in optimization time, and up to 70% reduction in hardware costs”.
The second big news is their Size Optimization for apparel. The traditional model of supplier pre-pack assortments of so many items per size, color, and style, has historically resulted in the all too common double curse of shortage on some sizes (bad customer experience, lost sales) and too much of others (and therefore, margin reduction from excessive markdown). New supplier and supply chain capabilities make it feasible to customize pre-pack to store level – but it doesn’t make any difference if you don’t know WHAT size assortment makes sense for each location. Their analytics package purportedly solves this problem. They cite great success with this already at Kohl’s, Charming Shoppes (especially Lane Bryant banner), Tilly’s, Aéropostale, and Wet Seal. Kohl’s Kevin Mansell (Chairman, CEO) cited size optimization initiatives as a key driver in their recent 3% improvement in in-stock levels. www.sas.com
Soft Solutions is a 20+ year old global provider of merchandising software for multi-format, multi-divisional and multi-national retailers. Their current product, version 7, released last year boasts a new interface that leverages their “Web flow” approach to enhance navigation, improve speed and capacity, and employ self-learning functionality – another key approach to optimization solutions. They play heavily in European markets, but also count large U.S. customers like CVS Caremark.
As with any merchandising technology, developing the right solution requirements up front is critical to success. For Tier 1 customers, Soft Solutions uses a certified third party provider to document requirements and make sure they are solid. The third party is paid by the retailer; but Soft Solutions offers a “cost offset”.
According to Dr. Fady Garabet, General Manager North American Operations, their forecasting includes halo, similarity and cannibalization analytics. CVS, for example, forecasts a full 52 weeks ahead. They believe their forecasting functionality is one of their great strengths. www.softsolutions.fr/eng/
Formed in 2000, 1010data’s solutions were developed for large scale financial data systems, and their leadership is heavily financial in their pedigrees. They strongly position their differentiation on merits and enablement of “database transparency”, contending that typical solutions, with the rules, filtering, and pre-processing of data for efficient access effectively lose some of what matters. “The problem with optimization solutions,” says VP Jim Mattecheck, “is that they have to extract into separate systems, which bottlenecks processing – they resolve this with averages, which obscures insights.”
Jim admits they “don’t have a formal optimization solution, per se, today”, but contends that retailers can benefit from the ability to do quicker analysis on questions that arise during their daily decisions, rather than waiting for long overnight processing. www.1010data.com
BrandTech News’ view
The exhibitors’ directory for the NRF show listed almost 40 companies with solutions in price and promotion optimization, so the above are but a few. This leaves potential new users (or disgruntled current users) with a quandary: how to pick the Optimal Optimizing solution.
Focusing as we do on the intersection of branding and technology, we suggest starting with the brand point of view. How you price your merchandise is a big part of your brand image, and a critical aspect of how consumers experience your brand. Some optimizers refer to the ‘price image’ as a way of characterizing this. Retailers need to begin with three questions on this. What do you think your price image is? What do your shopper segments think it is? What do you want them to think?
There are lots of technology issues to be considered, starting with a fundamental platform questions. Do you want Software as a Service (SaaS), premise-based software (installed on your servers), or cloud-based? What optimization modules and features will be most important to supporting your business objectives? Even more than the technology, understand the science behind it. Yes, there is real science at the heart of this, and a solid choice in solid science will help ensure solid results.
Most importantly, focus on value. The whole point of optimizing is a balancing act, pleasing consumers who will reward you with sales (and, if you’re good at it, margin), and pleasing shareholders who will reward you with market capitalization (and, if you’re good at it, performance bonuses). JTS