Introduction
Bob Stetzel, Vice President of Information Technology (IT) at Vermont Teddy Bear (VTB), walked a tranquil path from his car to his Shelburne, Vermont office early one morning in mid-February 2010. The landscape outside his office, and the White Mountains beyond, were blanketed in a coating of fresh snow. Just a few days before, the scene was not tranquil at all; a small army of nearly 2000 temporary employees had descended on the company’s multi-building campus to help process and pack gifts ordered by tens of thousands of customers for delivery to their sweethearts for Valentine’s Day. Bob and his seven person IT organization had worked feverishly behind the scenes, ensuring that the company’s information systems could handle the surge in orders for pajamas, custom teddy bears, flowers and other gifts, placed via telephone, mail-order, and the Web. There were a few tense moments when the system - comprising a mix of homegrown and packaged applications from a variety of vendors, and knit together with middleware - occasionally ‘paused’ when its capacity was strained. Fortunately, his team - veterans of past Valentine’s Day ‘peak experiences’ - helped patch things together and ensured that nearly all orders were processed and delivered on time. Recognizing that customer retention was an important goal, Stetzel was relieved that most customers were happy with the service they received during the Valentine’s rush.
Stetzel had been hired in November 2009 - just in time for a Christmas rush which included several tense moments as the systems struggled to handle a surge in orders. He hoped that before winter 2011 rolled around, his team could tame the complicated middleware and make progress toward an efficient, well-organized enterprise IT architecture that could serve as a robust platform for the company’s changing business requirements and support their long- term strategic and operational needs. Meanwhile, Stetzel knew he could expect another frenzy of activity in early May, when orders would peak just before Mother’s Day (second Sunday in May). Software bugs that had been identified during the Valentine’s rush needed to be fixed by members of his team before this next peak experience. Stetzel also wanted to carve out time in March and April to sort out his priorities for the IT organization so that the relatively calm days between June and November (when there were no major holidays to cause orders to suddenly surge) could be put to good use modernizing the aging and complex systems on which the company relied.
What project to do first? Should he oversee the selection of new enterprise software to replace the accounting systems responsible for order-entry, sales, and inventory management? This would replace a lot of problematic middleware, but at a high cost. Would it be better to focus on building or buying new supply-chain software to enhance operational capabilities, enabling raw materials and finished goods to be more efficiently procured from vendors around the world? Should the company acquire or build a Customer Relationship Management (CRM) package that would help them serve customers well across multiple product lines (bears, pajamas, flowers, other gifts) and channels (stores, mail-order, web, telephone)? Stetzel also wanted to build a new data warehouse and beef up the business analyfics capabilities that were needed for effective marketing. Clearly, his team could not possibly accomplish all these things in one 6-month period. Besides, the company’s cash reserves were quite limited in this tough economic environment. Difficult choices would need to be made soon, and a plan devised to get the job done.
As Bob Stetzel mulled over these concerns, CEO John Gilbert tapped on his door. Do you have a moment, Bob?
I learned some interesting things at last week’s Toy Fair (American International Toy Fair, held in New York City) that might affect our business - and perhaps your information systems.
Vennont Teddy Bea Company background
1981: VTB was founded by John Sortino, who sold teddy bears from a pushcart in a Burlington, Vermont open-air mall. The company nearly went bankrupt around 1990, but recovered when Sortino introduced a ‘Bear-Gram’ service, promoted via radio advertisements in the New York City area. Customers (mostly men buying for wives or girlfriends) phoned 1-800-829-BEAR to order a ‘persona- lized’ bear (choosing from several colors of bears and about 100 costumes such as tutus, wedding gowns, fire fighter and doctor or nurse outfits). The bear was shipped in a decorated hatbox with ‘air holes’ and a note from a ‘Bear Counselor.’ The market response to this promotion was impressive; revenues grew from less than $2 million in 1990 to $17 million in 1993, allowing VTB to raise $10 million in an initial public offering and earning it a ranking of number 21 in Inc. Magazine’s listing of American’s fastest-growing public companies.
Despite the success of the Bear-Gram service, numerous challenges threatened the company’s survival. Although radio advertising and a toll-free phone number generated lots of orders for teddy bears as gifts for Valentine’s and Mother’s Day (80% were purchased by adults for other adults), it was less cost-effective at other times of the year. In an attempt to induce adults to buy teddy bears for children throughout the year, the company began to sell through high-end toy stores such as FAO Schwarz, department stores such as Bloomingdales, and more than 200 shops. VTB also opened cornpany owned stores in New York City and Freeport Maine.
Unfortunately, these expensive moves combined with construction of a new company headquarters in Shelburne Vermont, entailed high expenses (leasing space for company stores cost $600,000 a year), which were not sufficiently offset by sales.
1997-1999: A new CEO, Elisabeth B Robert, closed the company stores (except the factory store in Shelburne) and launched a ‘Make-a-Friend-for-Life’ venture with retailer Zany Brainy: in-store machines assembled and stuffed personalized bears (similar to Build-a-Bear Workshops, a competitor). 1999 sales reached $21.5 million.
2001-2003: 2001 sales reached $37 million. Retail partner Zany Brain filed for bankruptcy, but Robert was not worried by this development, since We’re not a retailer; we’re in the gift delivery service business (Helmich, 2002). She re-focused VTB on direct marketing via telephone and website (55% of Bear-Gram orders originated online, 44% by phone). Producing about 450,000 bears per year in Vermont, VTB continued to target both children and adults. About 65% of orders were from men, with 30% of annual sales for Valentine’s Day (February 14). Their biggest customer was ‘Late Jack,’ an internal moniker describing men who ordered bears at the last minute instead of buying flowers or chocolates for girlfriends, wives, or mothers. To increase sales to women and to generate orders at other times of the year, VTB developed Gift Bag Boutique (handbags and baskets of food, accessories, and/or pajamas). In 2003, the company acquired high-end florist Calyx & Corolla (renamed Calyx Flowers). To increase sales to women and to generate orders at other times of the year, VTB developed Gift Bag Boutique (handbags and baskets of food, accessories, and/or pajamas). In 2003, the company acquired high-end florist Calyx & Corolla (renamed Calyx Flowers).
2005-2008: VTB reverted to private ownership. Revenues reached $66 million (Bears led at $31 million, followed by Flowers at $17.5 million and Pajamas at $14.5 million. Corporate sales, TastyGram, and other sources contributed the rest). Pressures intensified from strong competitors like 1-800-Flowers (which also sells gifts). The marketing VP expressed concern about demographic trends such as a drop in the number of young adult males (Exhibit 1). VTB dropped several product lines. Three remained: Bears, PajamaGrams, and Calyx Flowers.
Right after Valentine’s Day, 2008, 15 employees were laid off. In September Elizabeth Robert resigned, a month before an October stock market crash that reverberated worldwide. Interim CEO William York, a veteran of LL Bean and Dell, served during a time of further retrenchment. 2008 sales totaled about $75 million.
2009: A January press release announced VTB laid of 35 more employees. VTB retained about 200 employees; during three peak seasons (Christmas, Valentine’s Day, Mother’s Day) temporary employees were hired.
In March 2009, John Gilbert joined VTB as its new CEO, having previously served as Chief Marketing Officer for TJIX Companies and in marketing positions at Dunkin Donuts, Pepsi, Coca Cola, and elsewhere. In a press reléase announcing Gilbert’s appointment as ‘Chief Bear Officer’ Board Chairman Bob Crowley said:
An expert at building great consumer product businesses and brands, John is a natural choice to help lead Vermont Teddy Bear’s next chapter of growth and innovation.
Industy background (plush bys, pajanas. flowers)
In 2010, VTB relied on sales across its three separately managed brands, competing in three industries (see also Exhibit 2): * Teddy Bears (plush toys): A customer could purchase a stock teddy bear or could custom design one by selecting the color and outfit from a menu of options. At peak times, most purchases were for adult gifts; the rest of the year, purchases were mainly for children. All bears were produced on the Shelburne campus; fabric and other raw materials were purchased from vendors worldwide. * PajamaGrams (apparel/gifts): Production of pajamas was outsourced to low-cost overseas factories, so lead times were generally about 6 weeks. Thus, if an item suddenly developed a surge of demand, the company might not be able to fulfill orders in time for key peak holidays. * Calyx Flowers (flowers/gifts): This company, which began life as Calyx & Corolla, offered high-end floral arrangements for discerning customers who appreciated the long ‘vase life’ that was achieved because the company dealt directly with a small number of flower growers worldwide, eliminating the traditional middle- men who took some of the margin and delayed delivery of fresh-cut flowers to the end customer.
Teddy bears (Exhibit 3) competed in the plush toys segment of the playthings industry. Global toy sales were expected to top $80.3 billion by 2012 (www.npd.com/press/releases/press_090729.html) US customers accounted for almost 28% of toy sales worldwide; in the US, most toys were sold through ‘big box’ discounters (especially Wal-Mart), specialty retailers and the Internet. A large percentage of sales took place in the December holiday season. Strategies for generating sales at other times included offering collectibles, which customers purchased throughout the year (the Beanie Babies craze in the nineties is one example). Another tactic was to promote complementary accessories (e.g., clothing and other accessories for Barbie or American Girl dolls). A third tactic was to sign licensing deals with television and movie studios, game makers and others. US based Mattel and Hasbro, each with more than $5 billion in 2009 revenues, were the largest toy companies in the world. China, the world’s toy factory, produced 77% of 2008 US toy imports - about 20 million toys. Large companies tended to outsource production to Chinese contractors for toys with unpredictable forecasts (i.e., this year’s fad), and to use their own factories (often located in China) to produce classic toys (those with more predictable demand). The US Consumer Product Safety Improvement Act restricted the use of lead paint, phthalates and other harmful materials in children’s products, and mandated product testing and documentation requirements. For example, to support product recalls, every component of every toy must be traceable to its source.
PajamaGmms (Exhibit 4) competed in both the general gifts category and global apparel industry. China was a major producer of clothing; between 2002 and 2008 total imports of men’s clothing from China increased by almost 14%, while Mexican imports fell more than 6%. Most pajamas were sold through department stores (e.g., J.C. Penney, Sears, Macy’s), mass merchandisers (e.g., Target, Wal-Mart), and other outlets (e.g., Baby Gap, Victoria’s Secret). Although the global economic crisis continued to exert pressure on this industry, the VTB marketing VP reported that other trends were favorable. For example, increasing numbers of women were delaying having children, leading to increases in disposable incomes and demand for premium apparel products.
Calyx flowers (Exhibit 5) competed with thousands of competitors in the highly fragmented florist industry. The top four firms accounted for less than 4% of industry revenues. Mass merchandisers and large supermarket chains also competed in this segment, offering lower prices but limited selection. Since 2005, sales had steadily declined in this industry (IBIS, 2010). Most florists purchased flowers from wholesalers or jobbers, who purchased from growers. Online or catalog florists enjoyed better margins than their bricks and mortar competitors. Calyx Flowers (and some other direct marketers) purchased directly from growers and coordinated with shippers such as UPS to ensure that highly perishable fresh cut flowers would be refrigerated and given special care en route, to protect against wastage and unhappy customers.
Bob Sbtzel and the IT organization
Before joining VTB, Bob Stetzel worked for 12 years as a database analyst and administrator at the Holstein Association in Vermont, 2 years in a similar role at Fleet Bank, and 10 years as Vice President of Business Information Services at Yankee Candle Company. During his career Stetzel had overseen database administration, software development for traditional and eCommerce website projects, and implemented enterprise applications for content management, CRM, business intelligence, and supply chain management. The timing of his hiring - November 2009 - meant that he had to quickly take stock of VTB’s IT architecture and the capabilities of his small organization, since the holiday season would peak between Thanksgiving (fourth Thursday in November) and Christmas (December 25). Stetzel learned that five IT employees had been laid off during the 2008 - 2009 downsizing, leaving just seven IT employees under his management. He was pleased to discover that most of these individuals had excellent expertise, although some in narrow specialties. But, they spent too much time troubleshooting, leaving little time to develop new applications or implement new versions of packaged software. He was concerned that, were certain individuals to leave, they would walk away with important knowledge about the company’s IT applications or infrastructure, which had not yet been captured or codified.
His staff briefed Stetzel on initiatives that were underway, including one large project that was ‘a little runaway,’ in his opinion. ‘The application development practices and processes were not in the best of shape, to say it nicely.’ He declared a moratorium on that project and deferred all new development until the company could get through the two winter peaks (Christmas and Valentine’s Day) and he could get a handle on priorities.
An ideal IT architecture document would describe the network configuration, servers and major enterprise applications, and data flows within and across these (especially those applications involving transactions that affect financial statements). Stetzel knew that organizations that depend heavily on IT for strategically important operations need a disciplined change control process to ensure orderly procedures for requesting system modifications, developing and testing those in a separate environment, certifying them and introducing them to the production environment (‘hardening’). He was disappointed but not surprised that an incomplete IT architecture document existed at VTB, and that change controls were minimal. This was because the formal IT organization had not instituted such processes nor prepared such a document, and also because many ‘shadow’ applications had been developed by users to fill various needs when the IS organization lacked resources to help. Some of these applications were connected via middleware to the core enterprise applications that were under the control of the IT organization, while others were run independently. At a recent meeting, Stetzel had mentioned to John Gilbert:
Generally what happens is that somebody will call my staff and request some urgently-needed new functionality. The IT person who gets the call will do his best to get it done, but that might mean that a more important task is neglected. This is a typical problem for companies of our size, but it would be better to have a more organized approach to triaging and prioritizing these support calls.
Stetzel needed a complete, current catalog of all computing hardware and shared applications in use. Realizing that a lot of the necessary information resided in the heads of ‘siloed’ individuals (‘one person understands a particular application, but nobody else does’) he put together a team of several IS employees who had been with the company for more than 4 years, to reveal all this hidden information. This discovery process yielded good and bad news. The core transaction - processing infrastructure was ‘pretty strong for meeting the operational needs of a company this size; pretty forward- thinking,’ in Stetzel’s opinion. There were good practices in password management and recovery mechanisms (if a network element failed, built-in redundancy allowed the sales order systems to continue running so that orders did not get lost, for example). And by hosting their ecommerce applications at collocation facilities outside Vermont, they were able to get a more advantageous position on the national Internet backbone.
A ‘best-of-breed’ approach had guided the development and acquisition of enterprise applications that supported operations, meaning that instead of buying a single Enterpise Resource Planning (ERP) package whose separate elements might be adequate but not optimal, the company had purchased separate packages that offered closer to optimal functionality; these connected to each other via middleware. For example, software from Avenis, now named Circle Commerce, based on a UniVerse database, was utilized for order management and fulfillment. Bob Stetzel was not familiar with UniVerse, an IBM product that in October 2009 had been sold to Rocket Software; Stetzel’s extensive database experience was with other relational packages such as Informix and Oracle. A package from IndustriOS was utilized for managing inventory, and Platinum (Sage PFW - Platinum for Windows) handled accounting applications. VTB’s one remaining retail store utilized FusionRetail. Some of these applications were current (whenever a vendor released an upgrade, it was installed); others were several versions behind schedule. Beyond this core set of transaction processing applications there was a myriad of small one-off applications and middleware, based on various languages and tools and chosen more out of individual preference than an overarching plan. Stetzel observed that a programmer might say, ‘Hey, I know Perl (a popular scripting language), I think I’ll write this in Perl’ or PHP (another scripting language) or ‘I’ll use mySQL’ (an open-source database product supported by Sun Microsystems, a division of Oracle Corporation) or ‘I prefer Access’ (a Microsoft database product).
Soon after Stetzel was hired, CEO John Gilbert told him: ‘You need to figure out what the middleware is and how to manage it; it’s a black box.’ Now, having gone through his ‘cataloging’ exercise, Stetzel concluded that middleware was, indeed ‘the biggest challenge that I have walked into’. He realized that the middleware was not just being used to translate data among disparate applications, but itself contained a great deal of business logic, most of which was undocumented. Stetzel reflected on the ramifications of the ‘black box’:
We have three different brands and four different channels (retail, mail, phone, web), each supported by different software, creating a spider web of connectivity.
The application developers no longer work here. This black box is holding us up from doing all sorts of things to support both our traditional and newer eCommerce processes. We don’t yet know how it all works, nor how to pull without breaking it.
In his three-plus months as Chief Information Officer, Stetzel felt he’d made some progress in determining ‘what we have, where it runs, what are the outstanding issues.’ He hired two employees (total IT staff now nine people; see Exhibit 6) and authorized his team to make small adjustments aimed at keeping the systems functional and the processes they supported operational. In February Stetzel told a colleague, ‘Right now, we can ride it like a bucking bronco, but we certainly haven't tamed it yet.’ Meanwhile, users were continuing to request enhancements, most of which he could not deliver. He explained that it was imperative to first tame that bucking bronco: ‘Even though we're a small company, we have a lot of systems, and they haven't been managed effectively over time. We’re paying the price for that.’
John Gilbert returns from Toy Fair
CEO John Gilbert occasionally met with Stetzel to discuss planned new products and other business initiatives, in light of Stetzel’s plans to build a more robust and adaptable IT platform. On the winter morning when Gilbert stopped by, it was to talk about the annual American International Toy Fair that Gilbert had attended, along with 1100 other exhibitors promoting more than 100,000 products to 10,000 retail buyers. ‘While you focused on making Valentine’s Day as ‘fulfilling’ as possible for our customers, I was learning about the latest toy trends and connecting with potential partners,’ the CEO said. ‘Was it all about electronics? asked Stetzel. Gilbert replied, ‘There certainly were a lot of high-tech toys and games.’
I saw lots of interest in classic toys as well. Board games like Monopoly and Scrabble are going strong, and buyers showed interest in teddy bears. In this down economy, nostalgia is big; people take comfort in simple pleasures and filmily focus; we need to capitalize on that mood. It’s clear that Vermont Teddy Bear is a well known brand; maybe we should bundle a bear with pajamas.
PajamaGram’s Hoodie Footie product (a ‘onesie’ for women), introduced in 2009 was so popular that they had run short during the Valentine’s rush. Stetzel noted that with better data integration and improved analytic tools it would be easier to coordinate marketing, forecasting and production across the product lines to implement a family-focused ‘comfort’ strategy (‘pajama sets for the whole family”). Gilbert wondered if it was possible to ‘stretch’ such a promotion from Christmas through Mother’s Day, smoothing the peaks. He also wondered whether it might be possible to cross-sell pajamas or related products and flowers such as when flowers are purchased for women convalescing from childbirth or surgery. Calyx Flowers was largely independent from the others, so a lot of integration would be required to carry out such a strategy.
With the discussion of cross-selling, the conversation turned to CRM software. Stetzel explained the challenge there:
CRM works great when your customers buy products for themselves, over multiple channels. LL Bean uses it to ensure that they know who you are and your order history, no matter how you contact them - in store, mail, phone, Web. Most of our customers buy products as gifts, sometimes for both their girlfriend and mothers, which are two very different market segments. CRM isn’t as helpful in that scenario.
They discussed these ideas for a few minutes and then Gilbert commented that the Toy of the Year (TOTY) awards, conferred at the Toy Show, were quite interesting. ‘An ultra-simple toy called Bilibo earned Best Preschool toy (in a tie with a Tonka talking truck). It’s just a colorful little scoop; even easier to produce than our bears! Who’d have thought? He showed Stetzel (www.toyassociation.org), where an advertisement stated:
Bilibo is a new kind of toy. Requiring neither batteries nor complicated instructions, it’s powered purely by a child’s imagination. In their hands these open ended toys come alive and turn into anything and everything.
‘Were there a lot of other plush toys at the Fair, beyond bears?’ Stetzel asked. Gilbert said that Webkinz, which won a TOTY award a few years earlier, had launched a Webkinz Junior product and website, aimed at 3-6 year olds. In both the regular and junior versions, each Webkinz toy had an embedded chip containing a ‘secret code’ to activate an account on a website. He showed Stetzel (www.webkinz.com). ‘I had no idea that children that young would respond to a fantasy world on a website,’ Stetzel remarked. He added, ‘To emulate that strategy, we’d have to hire quite a few people to help with website design, not to mention creative types to generate content. We would also need a strong content-management system.’ The overall Toy of the Year had been awarded to Zhu Zhu Pets, inexpensive but ‘smart’ collectibles:
Zhu Zhu Pets drive around in little cars, zip up ramps and spin down slides, run in their hamster wheels, and get into all sorts of crazy situations. Their artificial intelligence even allows them to know what room of their hamster habitat they are in. They make toilet-flushing or teeth-brushing sounds when they enter the bathroom and sleeping noises and alarm-clock sounds when they go into their bedroom. (wwwzhuzhupets.com)
‘I doubt we want to produce toys that require artificial intelligence,’ said Gilbert. ‘But I do admire the Zhu Zhu Pets collectibles idea; both children and adults buy these things throughout the year. Clearly there’s a lot to think about.’ Stetzel agreed, and informed Gilbert that soon he would present him with a document laying out his plan for fixing or replacing existing systems and investing in new ones. ‘I look forward to that discussion,’ said Gilbert, adding:
We’re an ecommerce business, so the stability and reliability of our ecommerce platform, especially at peak times, is paramount. By now you probably realize our supply chain management could improve; we’ve had stuff arrive late or be poor quality. There’s also plenty of room for improvement in the systems that support our marketing - once we know who we want to talk to and what we want to say. Our current customer is actually a 52 year old guy, yet our creatives have been trying to talk to younger men. We’ve been advertising the wrong message for a long time, and it’s taking a toll on our peak occasion sales.
Planning for sung
After John Gilbert left his office, Bob Stetzel reflected on the days ahead. The snow outside reflected brightly. It wouldn’t be long before mud season descended on Vermont - Stetzel’s least favorite time, but a precursor to spring. which - despite calendars that suggest spring starts in March - would arrive in May, right around Mother’s Day. Stetzel doubted that he would notice blooming flowers before this next ‘peak experience.’ He hoped that his team’s efforts to complete the task of documenting the existing IT architecture wouldn’t get mired in its own ‘mud;’ this would prevent him from moving forward with a compelling proposal to modernize and standardize systems to support an evolving business model. He could eliminate a lot of the complicated middleware and incompatible applications by investing in a full -featured ERP package, but it would come at a high cost and probably could not be completely implemented before winter 2011. Besides, Stetzel knew that the benefits from an ERP implementation usually don’t materialize for another year to year and a half, and that it can be a traumatic process for employees. ‘It would be nice to show tangible progress filster than that,’ he thought. Where to start? And, whatever path he were to choose, how to convince Iohn Gilbert and the Board of Directors to provide the necessary resources?
Notes
1 http:/Mwwnpdcomlpress/releaseslpressfim729.11tml. 2 Software companies and Open Source Software tools mentioned in the case include the following

Behrenoes
IBIS (2010). Florists in the US, [WWW docunnent] 29 lmuary, http'll wwjbisvnrldcornl. Heinrich, P. (3)02). Not your average bear, Business People-Vermont, 2nd
February. Shank. LB. (MG). The changing profile of the United Smes. Library of Coupes: Congresional Research Service, 10.32701, 5 May.
About he Armor:
Ianis L Gogan is Professor of Information and Process Management at Bentley University, and holds M.Ed, MBA, and DBA degrees from Harvard University. Dr. Gogan’s research focuses on IT-enabled innovation, inter-organiza- tional collaboration, and global strategic IT alignment. Her publications include teaching cases, which have been taught in US, European, Australian and Asian schools, and over 100 papers in conference proceedings (such as AMCIS, ECIS, ICIS, and HICSS) and journals such as Communica- tions of the Association for Information Systems, Electronic
Markets, Government Information Quarterly, Information Systems Frontiers, International journal of Electronic
Commerce, journal of Information Technology, journal of Management Information Systems, and The Information
Society.
Mark 0 Lewis is an assistant professor of Information and Process Management at Bentley University. His research focuses on the strategic use of information technology, enterprise transformation, and investigating patterns of collective action within and between organizations. His research has appeared or is forthcoming in journals such as MISQ-Executive, European journal of Information Systems, and the journal of Global Information Technology Manage- ment. Before entering academia, Mark worked with IBM
Global Services as a business analyst and as a research assistant with IBM's advanced Internet technology division.
He has worked on research projects sponsored by companies such as UPS, Daimler-Chrysler, Hewlett Packard,
Gartner, and SAP.