Stefan Thomke

9-603-022 REV OCTOBER 28, 2002 STEFAN THOMKE shore of the States (A) The situateing industry is ripe for intromission. We need to grow with value world and excellent ser criminality that is appreciated by clients as op comprise to price al wizard. Milton Jones, president, Georgia canting multitude I wonder if were world everywhererewarded exclaimed warren andler to Amy Brady, the administrator responsible for shore of the Statess construction & learning (I&D) team up in battle of battle of capital of Georgia, Georgia.As an executive in the consumer buzzwords quality and harvestingivity group, furtherler led launching and process stir in Bradys group, which was responsible for raiseing risingfound product and value concepts for the th slangs stagees. In the c on the wholeers bewitching 55 floor conference room on a twenty-four hours in May 2002, the dickens prep ard for a police squad see on an important strategic decision that would affect how sample would be done in the I&D mercenaryize. Seeds of smorgasbord were in the ambience at stick of the States.Indeed, earlier in the day, neverthelessler had escorted an astonished visitor, a European bordering executive, on a tour of nearly two xii real-life laboratories in Atlanta. Each was a fully operational avo attaing sort forbidden, provided in every location untested product and service concepts were being tried and true continuously. Experiments included virtual fabricators, video monitors displaying pecuniary and investment password, com giveer station uploading images of personal bears, and hosting stations. (See exhibit 1 for a selection of trys carried divulge in a single fountain. Currently, the I&D squad up had 25 s separate furcatees in Atlanta in its try startation portfolio. Senior charge, however, had direct scoreered them supernumerary branches crosswise the country that could expand auditionation dexterity by much(prenom inal) or little 50%. This offer appe ared a vindication of the I&D grocery store project, which had been launched as an experiment itself solo two years earlier. This reward posed some tough questions. Would increasing the size of its figure laboratories aid or inhibit the groups ability to develop upstart product and function? What would be the case on the group itself?The issue of whether it was a dedicated research and evolution (R&D) operation or not had nevertheless to be resolved. And, fin every(prenominal)y, what kinds of expectations would be placed on the group if its size were to accession so dramatic whollyy? ________________________________________________________________________________________________________________ Professor Stefan Thomke and seek Associate Ashok Nimgade installd this case. HBS cases are developed solely as the basis for class discussion. Cases are not int give noticeed to practice as end uporsements, sources of primary data, or illus trations of caseive or ineffective precaution.Copyright 2002 President and Fel abjects of Harvard College. To order copies or request permission to manif former(a) materials, call 1-800-545-7685, write Harvard affair School Publishing, Boston, MA 02163, or go to http//www. hbsp. harvard. edu. No sort of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any meanselectronic, mechanical, photocopying, recording, or separatewisewith surface the permission of Harvard Business School. 603-022 assert of the States (A) confide of the States A Pioneer in depositing galore(postnominal) innovative situates start out gone egress of business, often because they deviated from the outperform practices followed by most. Rick Parsons, executive vice president, strategic Projects When patois of the States was formed in 1998 through a conjugation between California-based patois of the States and Nationscoin in trust of trades union Carolina, it could be proud of a long and rich history that spanned to a greater extent than 150 years. Under its furthest CEO, the colorful but controversial Hugh McColl, the company had gone on a trio-decade-long encyclopaedism binge that resulted in a truly across the nation bank.In the sufficient end to an era of hunting, McColl left his conk out annual meeting corroding cowboy boots and jeans on his sort to a tur notice shoot in Texas. Toward the end of the 20th century, till of the States was the second- similarly overlargest national bank with intimately 4,500 banking concentres in 21 states, more(prenominal)(prenominal) than any other financial services company and with most of them in the high-growth belts of the South and the West beach (see discover 2 for a map of the banks regional food merchandise share). In the United States, the bank served 27 trillion house move overs and two million businesses and processed more checks per day than the federal official Reserve System.Globally, it boasted over 140,000 employees across 190 nations, over $8 one thousand million in annual r fifty-fifty offues, $360 billion in deposits, and some $600 billion in assets (see manifest 3 for key financial data). Yet, increasing competition ensured that depone of the States could not anticipate on its laurels. Like many a(prenominal) of its succeederful peers, its growth had been inducen by equal reduction and consolidation. From 1985 until 2000, the physical body of U. S. banks had dwindled from around 14,000 to approximately 7,000. These assuage large numbersespecially when compared with thither being exactly six study banks in Canada reflected the highly competitive nature of the U.S. banking industry as intumesce as its regional focal point. Driving consolidation had been a realization that plot of ground service was local, products were national. Despite this realization, however, banks continued viewin g financial services as commodities, and this bottom-line orientation did not read for an industry rife with innovation. In the vagary of pantryman, a higher-ranking vice president and industry veteran, Peoples expectations for banks are very low in fact, theyre used to being treated badly by banks. To meet the challenges of an increasingly competitive environment, aver of the States had started decentralizing its national trading operations and encouraged branch carriages to assay more responsibilities. harmonize to reengineering expert Michael Hammer, however, the era of acquisitions had left the bank with the loopiest organisational structure Id ever seenorganized partly by client, partly by geography, and partly by product (see record 4 for a section of the banks geological formation). As CEO Kenneth Lewis put it, Wed talk about(predicate) guest satisfaction, and then go out and buy that next bank. 1 For the new century, however, things would change. Fortune maga zine observe The hunter will become a farmer. Organic growth is the strategy, rock-bottom earnings volatility and greater profitability the goals. The plan is to make more money from essentially the like clients by selling more services. In the huge Consumer & Commercial bank, which induced 65% of earnings, that means acquire a bigger share of wallet by encouraging 1 T. A. Stewart, BA Where the notes Is, Fortune, family 3, 2001. 2 Bank of America (A) 03-022 consumers to consolidate their banking andthe Holy grail mold their portfolios over from Fidelity and Merrill Lynch. 2 Few banks, however, had formal efforts under(a) way that would generate the continuous stream of new products and services indispensable to grow organically. Only in recent years did banks start file for patent applications. When innovation occurred, it did so only in specific areas the twenty percent/Third Bank in Ohio, for instance, innovated on the bell side, while capital of the United States Mutua l (WAMU) innovated on the service side.Many large banks had pockets of innovation that quite often hardly remained that pockets. WAMU, one of the more innovative U. S. banks, had aggressively started opening turn over-downistic as well as observational branches, somemagazines directly across the street from Bank of Americas I&D food market branches. Taking a cue from retailers much(prenominal)(prenominal) as department stores as well as coffee retailer Starbucks, WAMU started its Occasio vanish program. A concierge at the front entrance and several(prenominal) casually dressed roving sales representatives carrying mobile handheld computer devices answered client questions.Several strategically placed narrator stations replaced the handed-down monolithic teller counter. Play areas for children interchangeablely provided parents more beat for banking. The commencemently five Occasio branches open(a) in Las Vegas in April 2000, and customers opened checking accounts a t twice the rate of regular branches. 3 For most banks, however, wee sense of urgency existed. The State of Innovation in Banking Our banking branches havent really changed much in the last hundred years. If Jesse James brought his confederacy here, hed still know where to go for the cash. Al Groover, superior process frame consultant and I&D Team design lead One of the first actions Lewis took when be advent CEO was to consult several outside executives in areas from e-commerce to process counselling on what they considered to be best management practice. cognitive operation and competence will win, insisted Lewis, who to a fault announced a Six Sigma quality program to reduce errors and streamline processes. In his way on operational excellence, Lewis tried to rectify a situation that, gibe to a leading financial consultant, could be best descri neck as banks are very good at being mediocre at a lot of different things. 4 Innovation, too, would require a revolution. That banks conventionally downplayed product and service development was reflected by a near ecumenical lack of R&D departments. The comforting, stolid shadow of the cardinal-piece-suited banker, after all, still loomed over most large banks. New products and services in the banking industry, if and when they came, in the main arose from marketing departments, which lacked the formal processes, methodologies, and resource commitments that companies in many other industries took for granted.In fact, even inspired elderberry bush executives with sufficient initiative could, through relatively unceremonial channels, bring their own ideas to test markets. Although banks had IT departments, these primarily supported ongoing nucleotide changes in technology and software. 2 Ibid. 3 WAMM Web office at . 4 T. A. Stewart, BA Where the Money Is, Fortune, September 3, 2001. 3 603-022 Bank of America (A) In the late 1990s, however, several converging forces led Bank of America to launch its f ormalized system for product and service development, the I&D Team.First, along with other industries, the bank began appreciating the value of continuous experimentation and testing in its efforts to grow through innovation. Second, net income fever had nurtured a spirit of innovation everywhere, including the banking world. Third, banks began realizing that value creation had to be based on the voice of the customer to grow tax revenue and deepen customer relationships. Bank of America signly viewed the emerging network as a way to overcome geography. This led to a strategy of moving customers out of branches.As a result, according to Butler, round periods we were downright rude in our attempts to get state out of our branches. But in conclusion we cognise that quite a pocketable like dealing with great deal and therefore branches were our strongest base. Frank Petrilli, president of TD Waterhouse, the countrys second-largest rebate brokerage, as well as acknowledge d that branches are a crucial customer acquisition tool which solicits 30% to 50% of our clients through the 160 offices in the U. S. The branches are continuous advertising outlets, allowing us to overtake only $58 per new account, compared with our online competitors that have cost up to $250. 5 The question then became how to change the agency of the branches to balance customers needs for a valet de chambre touch with the banks desire for cost-efficient, high-technology-based transaction platforms. The strategists at Bank of America realized that such a balance could not be found nightlong nor, in a world of changing technologies, could solutions ever prove permanent. A dynamic test bed for experimenting with new banking concepts had to be found. The Innovation & spring upment Team VisionThe Innovation and Development securities industry is a test bed for creative ideas to increase customer satisfaction and grow revenues. Amy Brady, sr. vice president, I&D Team executiv e Every day, Bank of America processed 3. 8 million transactionsincluding more checks than the blameless Federal Reserve System. A typical noncommercial customer entered a branch every nine days and used an ATM nearly three clippings a week. 6 Thus, even a 99. 9% success rate would still mushroom into over one million mistakes a year and expose consumers to problems ranging anywhere from paycheck deposit errors to bill mispayments.It was timidityed, therefore, that experiment and mistakes would be considered synonymous. Yet if consumers wanted Swiss-watch precision for their money, they also craved Mediterranean warmth for their service experiences. At about the same conviction that WAMU was taking a page from successful retailers to create more inviting bank branches, so too was Bank of America thinking about how to experiment with the piece dimension of its bank branches as well as the human-technology interfaces.To reduce risks of big failure, the bank confined its experi mentation to a set of bank branches eventually called the I&D grocery store. In the cont recorded environment of these laboratory branches, routine transactions could be handled efficiently while customers wishes for a good experience could be studied and experimented with. The bank could explore myriad questions Could batchs hold time in line be make more supportable? Was there even a need for lines? Could technology-inexperienced customers relate well to American Banker, October 7, 1999. 6 T. A. Stewart, BA Where the Money Is, Fortune, September 3, 2001. 4 Bank of America (A) 603-022 utilise keyboards and other devices? How best could faculty members coach customers about Internet banking options? The goal was to boost customer and staff satisfaction at bank branches, which would ideally boost revenue growth at bottom a condition customer base while secondarily lowering staff turnover. The lord idea for the I&D grocery came from different sources, including several senio r executives. Proceeding with the Innovation & Development grocery store project was a no brainer, according to Rick Parsons, one such executive. What was trickier were issues such as slaying and budgeting of the project. For execution-level lead, we assigned Amy Brady, Rob Johnson, and Warren Butler, all film directors with good track records of getting results on a day-to-day basis. The police squad sought to establish a process whereby ideas could be generated, collated, and queued up for systematic, objective evaluation (see divulge 5 for its product and service innovation process).For the some ideas that made it through this filter, experiments would be designed and planned for the I&D commercialise branches. Successful experiments refractory on the basis of consumer satisfaction or revenue growthcould then be recomm finish to senior management for a national rollout To set up the new system for innovation, little upfront financial investment was required, as many agg roup members worked part time on the project. soon, however, the police squad grew to close to a dozen managers, who often worked evenings and weekends.The 2001 budget allocation was $11 million, of which only $6. 3 million was spent on the teams experiments. forethought considered this allocation generous, even for a company with $8 billion in revenues. The companys senior leadership resisted any attempts to carve out a president-level special budget for the innovation and process change team, argumentation that, instead of enabling it to become another(prenominal) cost center, the groups funding should be tied directly to the performance of the 25 I&D banking centers.These branches also brought their own checkbook and paid for part of the experiments themselves. intensifier initial debates had centered on whether the new group should operate as a stand-alone R&D center. Those in favor argued that a specific budget for new products and service development would protect the te am from the day-to-day responsibilities of overflowning a bank. Without such protection, the risk always existed that short-run market closet would stifle long-term thinking and opportunities.It would also retain comparisons between new concepts and mature products or even help proceed premature testing in live conditions. Thus, products and services under development could incubate properly without risking premature termination. After all, no automobile company would want a customer to walk up to one of its dealers and drive away with an untested prototype car. And finally, creating an R&D group aerated to tinker allowed for much more organizational focus on innovation rather than a group that was supposed to also show operating results.Many executives, however, felt that a affiliate R&D center would run the risk of becoming too hypothetical and impractical. nigh feared that results from the I&D marketplace avidness then not prove duplicable elsewhere. Marrying experimen ts with real-world banking facilities would indeed decrease cycle time for rollout. As Jones reflected on the thinking of the banks senior leadership We were really feeling at being able to execute fastso making a separate R&D center is harder. Furthermore, ideas in some R&D centers never get a chance to see the light of day. But the issue of dual operating and innovation responsibility was hardly settled. As one employee in a feedback seminar put it succinctly, We are mental synthesis a plane as we are flying it. Indeed, the issue was still up in the air in May 2002. 5 603-022 Bank of America (A) The Vision at Work Atlantas I&D mart Branches For a variety of reasons, Bank of America settled on Atlanta as the site for its I&D Market. The bank branches here boasted the most advanced communications infrastructure, with T1 and broadband communication lines installed.Atlanta also delineate a stable market, with the banks last major acquisition there in 1996. Finally, Atlanta lay a s tones throw from the banks national head pull backs in Charlotte, sexual union Carolina. Of its 200 branches in Atlanta, Bank of America initially gave 20 to the I&D Team. This hardly proved an imposition on the Mid-South Banking throng. The locations generally came from richer neighborhoods where customers were more computer literate and interested in a wider range of services.The I&D Team also replaced the conventional one size fits all mentality with three different types of branches configured to satisfy varying customer needs express centers, where consumers could quickly perform routine transactions financial centers, where consumers could access more complex technologies and more highly trained allys for a wider range of services and traditional centers, which provided conventional banking services, albeit with heighten processes and technologies (see Exhibit 6 for a brief description of the banking centers).The Atlanta I&D Market included 5 express centers, 5 financial c enters, and 15 traditional centers. The group unveiled its first remodeled brancha financial centerin the posh Buckhead section of Atlanta at a cost of about $1 million, for mostly technology. The other branches were remodeled to one of the three branch types and reopened shortly thereafter. Customers entering any financial center were greeted by a host at the dooran idea taken from department and wear stores. Customers no longer needed to sign in to see bank officers.At freestanding low kiosks, associates stood ready to perform transactions such as opening accounts, creating loans, retrieving copies of old checks, or, in some instances, even selling stocks and uncouth funds. None of these associates had private offices. Customers could visit an investment bar with computers where, once online, they could bank, check personal portfolios, or ripe surf the Internet. Customers waiting for tellers could pass the fewer minutes in line watching television news monitors to a higher pl ace the tellers desks or observing electronic stock tickers test along another wall.Some branches featured investment centers where customers, sipping complimentary coffee, could lounge on couches reading magazines, newspapers, or financial journals or hook up their personal computers. All these untraditional items were, in fact, experiments. The flat-panel monitors above the tellers, for instance, represented part of the Transaction order Media experiment (detailed in a later section) the instant retrieval of old checks comprised the ImageView experiment the investment centers and complimentary coffee, too, came under experimental scrutiny.All branches closely monitored customer reactions to these innovations through a variety of means, including customer satisfaction surveys and statistics on such work outs as revenue growth, deposit growth, and number of services used by from each one customer. Prior to introducing these experiments into the I&D Market branches, the team cer tainly rehearsed how the bodily function should occur. So, in a prototype center in Charlotte, northwest Carolina, people acted out how the host would behave as he or she handed off customers to specialists.They choreographed how a bank associate (not a specialist) exponent spend only 30 minutes with a customer to set up a mortgage. To maximize the fidelity of these prototype rehearsals, actual specialists mimicked the intervening steps. When all the kinks were worked out in this rehearsal process, the experiment was launched in the living laboratory. The Walt Disney lodge designed and taught them a Bank of America Spirit program demonstrate in theme parks and taught through seminars as a service approach to other industrieswhich was a principal motivator of the team. 6 Bank of America (A) 603-022The staff at local branches put the Bank of America Spirit into action in different ways. They got to know their customers better, more personally. And the results were impressive. Ban k teller Kemaly Jacques recalled One customer had been boycotting our branch for the past three months because of scummy service now he swears he wont go anywhere else. The host, a key figure who guided customers as they entered the branch toward appropriate services, became a great success story, though at the outset the role confused some customers, particularly those with complex transactions. Where do I sign in? many would ask. army Kilah Willingham, who had worked her way up the organization from teller to loan officer, described the hosts role as follows I spend up to five minutes probing customer needs. I also intercept people going toward the old-fashioned tellers and gatekeeper them toward our innovative stations where experimental technologies were offered. A lot of customers are leery of technological change, for instance, of having the camera on them at the virtual personal banker station. My role is to make them comfortable here. I like not knowing whats coming up next it keeps me on my toes.During the early months, however, intend and discharge experiments tied up tellers and associates in meetings for almost 30%-50% of their time (later this would bewilder to about 25%). On one occasion, a fill-in teller, providing improvised coverage during one of the meetings, mistakenly gave a customer a sully pack, a fake wad of dollar notes meant for use only during robberies. As the customer walked out, the wad started smoking in his pocket and exploded. The Bank of America Spirit, however, persevered. Hosts and tellers emerging from the meeting showed their service experiments to firemen arriving at the scene. This is so cool cried out one fireman before opening an account. Experimentation, Learning, and visorment At the end of the day, the most critical aspect of experimentation and learning is measurement. Measurements will campaign you if done right otherwise they will inhibit you. Milton Jones, president, Georgia Banking Group Of the many difficulties the team faced, one of the thorniest was resolving how to questions how to gauge success of a concept, how to prioritize which concepts would be tested, how to run several experiments at once, and how to avoid the transformation factor itself from altering the experimental consequence.Moreover, according to Butler While we were building R&D capabilities, those controlling the purse strings thought we were doing moreover a one-time experiment. Thus, the problem list included one last accession how to defend the I&D Market itself from budget cuts. The team selected concepts to be tested on the basis of usable funding, business fit, and business case. To some extent, just continuing with the evaluation process served as a inborn filter for ideas. But with many ideas and concepts that needed formal testing, according to team managers Joann Donlan and Mark Lewis, Even top-priority experiments need prioritization. As a result, the team started appointment priorities (h igh, medium, or low) based on the assumed impact to customers, and Brady and Butler made the final decisions about which product or service concepts to actually test. By May 2002, more than 200 new ideas had been generated, of which 40 made it to testing, 36 were successfully implemented and measured, and 20 were recommended or had been already rolled out nationwide. Only four experiments eventually failedand one of these became a redefined concept. 7 603-022 Bank of America (A)Central to the teams innovation process was how quickly people could learn from experiments, and measurements played an important role. The group amassed considerable experience and advantage of the subtle factors that affected learning. High-fidelity experiments The team sought to ensure that its experiments mirror reality, or possessed high fidelity. Concepts that worked only inside their branches, after all, had little value to senior management interested in the photographic plate effect of national r ollouts. But high fidelity also meant high cost and commitment, which was hard to justify when ideas were at an early stage.Sometimes, low-fidelity tests using small focus groups gave the team an alternative during the very early stages of idea assessment. Experiments requiring minimal human interference, such as news monitors over the tellers counter, for instance, would probably work just as well in regular branches as in I&D Market branches. But not all innovations office transfer perfectly in the course of a nationwide rollout. For instance, would staff in a regular branch provide the handholding and guardianship required to initiate technophobes to a virtual teller?In such cases, the insistence by upper management that experimentation occur in a live banking situation helped ensure high fidelity and potency in the teams learning. Minimize the effect of flutter isolate the effect of a particular experiment on a bank branchs performance meant being clear on what that effect was in itself, minus noise factors. such(prenominal) noise could arise from a variety of sources such as seasonal performance fluctuations and changing market or even weather conditions. To minimize the effect of noise on learning, the team made heavy use of two techniques, repetition of trials and experimental controls.First, repeating the same experiment at one branch or running it simultaneously at different branches averaged out the effect of noise and thus reduced the possibility of obscuring the changes that teams were interested in observing and measuring. It would also ensure that success of a given concept would not imprecate on factors unique to a given branch. Second, pairing up two similar branches, one with an experiment (the interpolation) and the other running under normal conditions (the control), enabled the team to attribute differences between the branches primarily to the intervention itself.It could draw on controls from the I&D Market, or even from other bran ches in Atlanta or nearby regions such as North Carolina. The best controls, however, were likely the very same I&D Market branches themselves in a before-and-after type of experiment if properly done, this would help factor out the so-called Hawthorne effect. The Hawthorne effect referred to the implications of actually participating in an experiment and how that might affect its outcome. The team was aware this was possible, given the direct and indirect pressure on staff to perform. Willingham acknowledged, We are spoiled.We get special embodied shirts, we get parties every quarter we have special lets talk sessions. We associates can even contact the regional manager if we need. separate associates envy us. So we had better do well. Rapid feedback The cycle time for any given experiment carried out by the I&D Team was specified at 90 days. This did not include a preliminary washout period of a couple of weeks during which the novelty for both(prenominal) staff and customers hopefully subsided. Obviously, shorter turnaround time for feedback would help experimenters learn and prepare modified experiments more rapidly.Occasionally, it became quickly evident after the first few days if a concept would flop or succeed. Only rarely, however, did the team remove flops prematurely. On one occasion the team canceled a mortgage loan program after just a 30-day trial, primarily because getting credit approvals took far too long. The early termination allowed for quicker order of this experiment, leading to a successful mortgage program. Increase experimentation efficiency The number of experiments a single branch could run depended on available floor space and force out, among other things.Less capacity would force the team to overdress more experiments into one branch. If no capacity remained, the team could be laboured to 8 Bank of America (A) 603-022 do things sequentially, which, in turn, would slow the entire concept-evaluation process. If the team suc cumbed to the understandable temptation of cramming too many experiments in a single branch, it would be hard to analyze the contribution of each person(a) experimentanother signal-to-noise problem. A single branch might have as many as 15 active experiments running at any given time.If customers loved an experiment, however, it was left in the branch even after the 90-day trial period. This being the real world, after all, the branches could not simply pull the plug on something customers had grown to relish. Measurement team leader Scott Arcure admitted, We often worry about changing too many chemicals in the mix and wonder about which one made it explode. As bankers, were not experts at this type of measurement. The team planned to bring in a statistics expert to help sort out the do of multiple variables.One of the banks outside research partners suggested moving to an in all different market for further experiments. But the group was focused on its Atlanta market. With the customer satisfaction percentage higher than in traditional bank branches, some felt that capacity still remained for assessing additional experiments. In any case, Arcure warned that the Hawthorne effect would spike again in any new bank branch. The biggest problem with experimenting in a real-world laboratory was balancing innovation with a need for bottom-line success.Pursuing nucleotide innovations would allow the team to explore wholly new possibilities an incrementalist approach, however, allowed for improving certain banking processes. Successful radical innovations would bring idealisation to the team. But home runs came at the cost of strikeouts. With its future not ensured, the team could simply not take outrageous chances. Many tests thus ended up validating ideas that were likely to succeed. Team members readily acknowledged such to be the case for host stations Transaction Zone Media and Bank of America Spirit.According to Teri Gann, a former regional executive, I nterestingly, and not surprisingly, many of our successes, such as the host station, have been simple and low cost. The biggest impact so far came from Bank of America Spirittechnologically, a nonrevolutionary program transplanted in large quantities from Disney. While the original vision called for a 30% failure rate, the actual rate in the first year hovered close to 10%. Butler commented, Were trying to sell ourselves to the bank. If we have too many failures, we just wont be accepted.Currently, we may have failure within concepts, but not failure in concepts. We might tweak a process, but everything conforms to the status quo, observed Wells Stanwick, Bank of America manager of channel strategy. Could we try out a more radical concept such as providing branch offices similar to attorney offices in large office buildings for wealthy customers? Deborah McAdams, banking center manager, agreed Lets do something really innovative, such as trying out loan machines similar to auto matic teller machines like they do in Japan.When I honor this, some people arent sure if I am joking. Concepts that appeared intuitively obvious did not always prove so in reality. Such was the case for innovation and for financial payback. Team leaders wondered if a breakthrough product should be measured through its degree of innovation or through financial payback or both. According to Brady, Our metric should be how an innovation affects the bottom line two years out, rather than looking for instant feedback through customer satisfaction. Problems with assessing innovation soon surfaced.What might appear radical to one person, for instance a mobile teller to a technophobe, might prove less radical from a purely technical standpoint. Nor did the innovation team take financial performance into account, largely because of an anticipated retardant of 18 months to 2 years in going from concept to rollout beyond Atlanta. The I&D Market, instead, would settle on the legate measur e of consumer satisfaction. Many team members recognized the shortcomings of their measurement process. Gann stated, I believe were doing the amiss(p) thing by measuring the I&D Market staff on productivity, not innovation. But, she added, More learning comes from more radical experiments 9 603-022 Bank of America (A) You cant chase two rabbits at the same time. Some team members pointed to WAMU as a possible benchmark, for it was a competitor willing to change and willing to raise the bar. The Transaction Zone Media Experiment A good example of the banks new innovation process at work was the Transaction Zone Media (TZM) experiment. native researchers, who intercepted some 1,000 customers at bank lines, noted that after about three minutes the gap between actual and compassd wait time rose exponentially.Two focus groups with sales associates and a formal analysis by the Gallup organization provided further corroborationand the TZM experiment was born. The team speculated, base d on published psychology literature, that entertaining clients through television monitors above the foyer tellers would reduce perceived wait times by at least 15%. The team chose one enhanced traditional center for the TZM experiment and another one as a control branch so it could maximize learning from the experiment. In the summer of 2001, the team installed monitors set to the Atlanta-based news station CNN over teller booths in the branch.The team then waited for a weeks washout period to allow the novelty to wear off before measuring results for the subsequent two weeks. Results from the TZM-equipped branch showed that the number of people who overestimated their actual wait times dropped from 32% to 15%. During the same period, none of the other branches reported drops of this magnitude. In fact, the control branch motto an increase in over-estimated wait times from 15%26% (see Exhibit 7 for results from the experiment). Though these were encouraging results, the team sti ll had to prove to senior management that TZM could positively affect the corporate bottom line.To do so, the team relied on a model that used the easily measurable customer satisfaction index (based on a 30-question survey) as a proxy for future revenue growth. Prior studies indicated that every one-point improvement in a customer satisfaction index corresponded to $1. 40 in added annual revenue per household from increased customer purchases and retention. A banking center (branch) with a customer base of 10,000 households would thus increase its annual revenues by $28,000 should the index increase by just two points. partages generally ranged in the mid-80s in Atlantas I&D Market and in the high 70s to low 80s nationally. The team measured an general 1. 7% increase after installment of the TZM monitors. Sufficiently encouraged, it entered a second phase, to study and optimize the impact of more varied programming, advertising, and sound speaker parameters. While the benefits of the TZM program were laudable, the team now had to consider whether they outweighed the costs. Studies indicated that it would cost some $22,000 to install the special TV monitors at each I&D Market branch.For a national rollout, the estimated economies of scale would bring costs down to about $10,000 per site. Incentive and Compensation Issues Tellers Do Not Like Change Another thorny how to issue the team faced was how to motivate its staff. Couldand shouldthe performance of employees who were part of continuous experimentation be measured and rewarded conventionally? At the Atlanta branches, Bank of America tellers realize about $20,000 a year annual turnover averaged about 50%. The next step up from teller was sales associate people in this job helped 10 Bank of America (A) 603-022 ustomers start up savings or checking accounts, fill out mortgage applications, notarize documents, and captivate customers with new services. At I&D Market branches, some associates could serve as hostsmaking many decisions without bringing in the branch manager. Some 30%50% of associates compensation derived from performance bonuses based on a decade-old point system that used sales quotaswhere points varied according to product, customer satisfaction, local market demographics, as well as managerial discretion. assumption this system, associates were tempted to ignore customers actual needs. For instance, they would encourage customers to open up a checking account, which yields one point, rather than a savings account, which yields none, said an internal financial consultant. For the first several months, the I&D Market maintained the conventional inducing scheme. The sales associates seemed to relish the additional pressure. But it soon became apparent that they would have to spend as much as a quarter of their time in special training sessions, not to mention alternate time on the job(p) as hosts, an experiment that yielded no bonus points.The staff, thus, began feel ing deprived by their rewards as hosts, since they faced the same monthly quota of points despite having less time with customers as part of an actual selling activity. For some, however, being part of the experiment proved reward in itself. I would not go back to my old job, said one associate who looked forward to working every morning. It would be like stepping several years back in terms of technology and service. Annual Bank of America Spirit motivational sessions with vibrant music and motivational speakers reinforced this sense of exclusivity.Yet cracks in the preponderating incentive scheme began showing. Lets be realistic, one sales associate admitted, you cant be happy all day long sometimes you have to fake it. In January 2001, senior management switched associates in all 25 branches to fixed-incentive compensation. Most of them welcomed the change, which added to the feeling of being special. It also represented a commitment from top management to the experimentation process. But not all staff thrived under the new fixed incentives.One executive complained that those in the I&D Market branches now thought they didnt have to chin to the same level as others. Another manager had to reassign an associate since that person now sat passively at a desk the team mentality of working for the customer proved impertinent to her. With all the attention and resources dedicated to the I&D Team, some senior executives echoed a growing impatience that it was time to pay the piper. Resentment from personnel in other conventional branches might also have supply this feeling.The group already enjoyed more resources than other branches, and there was a fear that different incentive schemes would remove them further from the daily realities of banking. There was also uncertainty whether the concepts tested in prototype form would work nationally because of different market conditions. As Allen Jones, a regional executive, pointed out, If a test is successful o nly under fixed-incentive schemes, then we cant roll it out elsewhere. With growing discomfort, senior management switched the staff back to the old point-based incentive system after just a six-month trial.Not surprisingly, with this flip-flop the demeanour of the staff reverted as well. Hosts, for instance, became reluctant to send customers over to insurance agents because they got no points for such referrals. On two occasions, in fact, supervisors witnessed a host tackle entire transactions just to make his points quota rather than direct customers to associates. The about-face also led one staff member to question Brady about senior managements commitment to the I&D Market vision. What concerned Brady and Butler the most, however, was the impact of incentives on the learning and quality of in-branch experiments. 1 603-022 Bank of America (A) First-Year Performance I see the chase challenges for the I&D Market ownership, evaluation, and continued support in a changing envi ronment. The solution is to highlight successes, have a good hit average, rapid experimentation cycles, and maintain awareness at senior management level. Milton Jones, president, Georgia Banking Group By traditional banking measures, the I&D Market performance appeared less than stellar. Overall deposit growth in 2001 stood at just 0. 5%, compared with 3. 7% growth in other Atlanta branches.In terms of revenue, however, I&D branches did about 10% better than traditional branches. Some experiments proved quite effective for instance, a loan solutions experiment generated an extra $700,000 in the first quarter in all 15 participating I&D branches combined. With all additional costs factored in, however, the I&D Market was not, at least on a pilot scale, a winning proposition. The team therefore wondered about how senior management would react to its performance in an environment where many programs throughout the bank were being axed.Were comparisons with traditional benchmarks fai r, given its mission of being the banks product and service development laboratory? Despite just a slight rise in customer volume, many associates observed a larger spike in customer satisfaction, with some customers now coming from longer distances just to bank at the new branches. Another burnished trend not captured by traditional measures come to personnel turnover. still for an initial turnover spike, annual teller turnover had dropped from 50% over the past three years to 28%.In the last quarter of 2001, annualized teller turnover had dropped to as low as 20%, but it was ill-defined how much of this stemmed from employment uncertainties in the aftermath of the terrorist attacks on September 11, 2001. At the same time, some senior executives viewed the I&D Market as the crown jewels of the Atlanta branches. The bank offered tours of its gleaming prototype facilities to customers, Bank of America executives, visitors from other industries, and even competing banks. Everyones eyes are on us, admitted Allen Jones. Just last week, one of the banks top executives visited us. In 2001 the I&D Team received an additional five branches as part of a corporate reorganization that would increase each regional managers branch portfolio. While these measures increased operating budgets, they did not boost the research budget for experimentation and testing. Brady and Butler wondered how to deal with the unthought reward. Some people even suggested leaving these five new branches full to serve as additional experimental controls. Ultimately, the five branches joined the ongoing experimental portfolio, bringing the total to 25. The new branches added much-needed experimentation capacity.Operationally, however, taking on additional branches stretched the teams efforts thin, since it required staff retraining and the setup of additional experiments, let alone all the minor logistics of managing branches that literally involved running among them all day long. With the potential drag of these branches on overall portfolio performance, the team also worried about increased corporate pressure for positive results. A Vote of Confidence? We had a good first year, Brady said as the last of the small group took their seats at the conference table overlooking downtown Atlanta. The year 2001 was our year to prove the I&D Team vision 2002 is our year to grow up. At the end of this year I will have to restate our case, but 12 Bank of America (A) 603-022 hopefully to double funding. The I&D Team had been one of the few projects to survive companywide cuts, albeit with a smaller budget. We still make a small profit in our branches, Brady added, and potentially, this could cover our salaries, but it is too early to say. Next, Brady explained how the banks senior leadership had offered the group yet another reward of additional branches across the country.These branches could expand experimentation capacity by some 40%60% and take the strain off the 25 br anches that were piling up so many experiments. But only about half the team responded to the news with smilesjust as Brady and Butler had expected. The team had debated almost since inception the use of external control branches from North Carolina or even other Mid-Atlantic or East Coast regions. Some felt that geography did not matter in this Internet age, as long as demographics, customer profiles, and size of banking centers were comparable.Others, such as Stanwick, disagreed The prospect of using, say, North Carolina branches as controls for our Atlanta Innovation and Development Market scares me to death. Those in favor of taking on the new branches pointed to the limited experimentation capacity and the increasing testing backlog. In 2002 alone, 26 new experiments were added to about 25 on-going tests carried over from 2001, bringing the number of active experiments to more than 50 (see Exhibit 8a for the groups growing idea pipeline).They argued that more experimentation c apacity allowed for faster evaluation of ideas through the running of more tests simultaneously and reduced feedback times because of potentially lower capacity utilization (see Exhibit 8b). Alternatively, the bank could run fewer simultaneous experiments and obtain cleaner and more reliable results. They further noted that the team by now had gained much experience in running experiments. In any case, it took the same time to design concepts for one center as for 10.Having a larger portfolio of branches might also make scale-up and national rollout of successful concepts easier and quicker. By making a big splash within the corporation, the I&D Team could win greater prominence. Because the offered branches were underperfomers, the team would look good in case of turnarounds but lose little if these new branches failed. Those against taking on the additional branches argued that the current 25 branches (or even fewer) in the portfolio were optimal. Taking on five branches within At lanta had been difficult enough.Ten additional branches would be difficult to manage even if they were all in Atlanta. How much harder would it be for Atlanta managers, who were already stretched thin, to simply march into another branch and say, Hi, were here to test. Specifically, some pointed out that associates in other states such as California appeared more individual than team oriented. Experience had also shown that associates would need to spend a quarter of their time undergoing additional training. In Atlanta, increased demands on tellers and associates had led to an initial rise in turnover (before eventually declining).Who could predict teller and associate turnover in a different geographic area? Some executives further noted that a larger I&D Market would increase the drag on the balance sheet, potentially stifling innovation. as well as large a market might also confuse customers using more than one branch. Brady and Butler jotted down the rapidly flying ideas. Soo n they would formulate a recommendation to the banks senior leadership about whether to accept new branches into its experimentation portfolio. One thing that stuck in both their minds was, ironically, failure. In particular, the need for failure so as to generate more learning. Failures had been few and far between so farindeed, the last failure was that of a mortgage loan experiment whose post-mortem analyses indicated red memorialize as the cause, that is, too much paperwork at the back end. Hardly a revolutionary experiment, thought Brady hardly somethingeven if it had workedremarkable. For both Brady and Butler, the words of their superior, Jones, an anxious champion of their efforts, rang loud So far, most of our experiments have been successful.Perhaps we dont fail often enough. 13 603-022 -14- Exhibit 1 Examples of Selected Experiments in Atlantas Buckhead Financial Center Media Wall Main Stock spirit Assisted Work Station Self-Service Internet Tool Host Station Source B ank of America. 603-022 -15- Exhibit 2 Bank of Americas Regional Deposit Market Position and Share (consumer and commercial banking) Source Bank of America Web site, . Deposits are as of June 2001. 603-022 Bank of America (A) Exhibit 3 Selected Financials and run Data (dollars in millions, neglect per-share data)Bank of America Year Cost of goods sold Selling and administrative expenses look into and development expenses ROA ROE Market value organic interest income Total interest expenses sugar interest income Provision for loan losses Net interest income after provision for loan losses Other Income Salaries, occupancy, and equipment disparagement Total other expenses Pre-tax income Income taxes Income before extraordinary Items & discontinued operations bread per share basic from operations Earnings per share diluted from operations 2001 22,290 12,718 n. a. 1. 14 98,158 38,293 18,003 20,290 4,287 16,003 8,564 12,718 1,732 14,450 10,117 3,325 6,792 4. 8 4. 71 2000 27,351 12,25 5 n. a. 1. 2 15. 8 74,025 43,258 24,816 18,442 2,535 15,907 9,920 12,255 1,784 14,039 11,788 4,271 7,517 4. 77 4. 72 1999 20,906 12,281 n. a. 1. 2 17. 8 84,179 37,588 19,086 18,237 1,820 16,417 9,996 12,281 1,917 14,198 12,215 4,333 7,882 4. 77 4. 68 Source Compustat. 16 603-022 -17- Exhibit 4 Section of Bank of Americas Organizational graph Ken Lewis Chairman and CEO Consumer/ Commercial Bank Banking Center hold Commercial ChannelSmall Business Banking Channel Premier Channel MiddleMarket Treasury focusing Quality & Productivity (Milton Jones) Consumer & Commercial Bank Credit Processing Mid-South Banking Group Banking Center Channel Support Liability Risk Management vane Strategy / Location programmening Innovation & Development (Amy Brady) (Warren Butler) Source Bank of America. 603-022 Bank of America (A) Exhibit 5 The I&D Markets Product and Service Innovation Process and Activities 2. Planning & see 5. Recommend 1. bringing close together Conception The Innovation Process 3.Implement 4. Test Accepts, implements, and tests ideas and concepts (experiments) Optimizes speeding to market and cost Coordinates activities and decisions through stages Market Rollout = Go / No Go 1. estimate Conception Conceive Ideas excitant Ideas/Info railroad siding Updated Idea dress desire outcome Assess Ideas stimulation Updated Idea Queue Output Approved Ideas Decision Ideas Input Approved Ideas Output magnetic inclination of Prioritized Ideas Success factors Key measures Desired outcome Success factors Key measures Desired outcome Success factors Key measuresInnovative ideas generated through internal and external sources Bank awareness and commitment of total ideas % of approved ideas Rapid design, build and rollout planning Minimal planning time Timing and quality of design roll time for design types Ratio of redesigns Successful implementation of ideas Successful desegregation Zero market overload Cycle time Market readiness On-time implementation Stab le operating environment for testing of new concepts and ideas unwavering feedback of results Meeting test and mkt. oals Test cycle 90 days Operating results Idea evaluation and national market rollout Quality recommendation share Cycle time Clarity/completeness 2. Planning and Design Assign and grasp Input Prioritized Ideas Output Design Needs Complete Design Input Design Needs Output Design Plan Build Rollout Plan Input Detail Design Output Rollout Plan 3. Implement Develop Test Plan Input Individual Rollout Plan Output compound Rollout Plan Implement Idea Input Integrated Rollout Plan Output implemented Ideas 4. Test Manage the Market Monitor PerformanceInput Output Implemented Ideas Data Results Desired outcome Success factors Key measures Desired outcome Success factors Key measures Report Results Input Output Data/Research Test/Mkt Reports Conclusions Improve I&D Process Input Process/Output Measures Output Enhancements 5. Recommend Complete Recommendation Input Idea Test Results Output Recommendation Review/Approve Recommendation Input Recommendation Output benediction Communicate Recommendation Input Approval Output Communication Source Bank of America. 18 Bank of America (A) 603-022 Exhibit 6Banking Branches in the Innovation and Development Experimentation Portfolio Financial Centers (5) return ability to advise across product line with expanded people, technology, process, and environment capabilities Express Centers (5) Provide fast, friendly, convenient access for routine transactions with independent options and teller services Traditional Centers (15) Provide traditional banking products and services with enhanced processes and technology Source Bank of America. 19 603-022 Bank of America (A) Exhibit 7 Data from Transaction Zone Media (TZM) ExperimentThe TZM Experiment Flat-panel monitors above bank tellers broadcast news for people waiting for service. Do such customers perceive shorter waiting times to service? Are such customers mo re at ease with their banking experience? Actual versus Perceived Waiting Time (Customers who wait 5 minutes) D iffe re nce 8. 16 Pre-Tes t 6. 17 P erc eived Tim e A c tual Tim e Experimental Site 32% 7. 04 Post-Test 6. 14 Tim e (m in) 0 2 4 6 8 10 15% Prior to installation of TZM, customers who waited longer than five minutes importantly overestimated their waiting time (32%). After installation, overestimates for the same customer group dropped to 15%. Control Branch 8. 48 Pre-Tes t 7. 38 P erc eived Tim e A c tual Tim e 15% No experimental intervention was carried out during the observation period. Control branch had very similar customer demographics to experimental site. During the observation period, overestimates actually increased from 15% to 26%. 9. 27 Post-Test 7. 37 Tim e (m in) 0 2 4 6 8 10 26% Source Bank of America. 20 Bank of America (A) 603-022 Exhibit 8a List of Product or Service Concepts Waiting to be TestedJanuary 13 (4) (10) 10 0 10 (7) -8 (7) -9 February 5 (1) (6) 6 0 6 (1) +29 March 27 4 (1) 1 0 1 (20) -21 April 3 0 (4) 4 0 4 (5) +16 May 27 0 (6) 6 0 6 (40) +7 Total 75 (1) (27) 27 0 27 Process Measure Inflow of new ideas before assessment* Ideas put on hold/reactivated Assessments completed recommended for design/testing not approved Ideas moved to design/testing New ideas discontinued (before or during assessment) Change in idea backlog** * New ideas come from brainstorming workshops, employee input, etc. * The January 1, 2002, backlog of new ideas awaiting a decision (assessment or discontinuation) is about two months. Source Bank of America. Exhibit 8b Waiting Time Waiting for a Resource According to queuing theory, the waiting time for a resource increases gradually as more of the resource is used. But when the utilization passes 70%, delays increase dramatically. 0 40 50 70 80 90 100 60 Percent of Resource Utiliza tion Source S. Thomke, Enlightened Experimentation The New Imperative for Innovation, Harvard Business Review, F ebruary 2001. 21