bagel sales double at host marriott using qfd · bagel sales double at host marriott ......

Bagel Sales Double at Host MarriottUsing Quality Function Deployment

Steve LampaBrand Executive

Marriott Hotels, Resorts, and Suites

Glenn H. MazurJapan Business Consultants, Ltd.

The Eighth Symposium on Quality Function DeploymentInternational Symposium on QFD ‘96

Novi, MichiganJune, 1996

The Eighth Symposium on Quality Function Deployment June 9-11, 1996

Japan Business Consultants, Ltd. values the spread of ideas.In this spirit, you have permission to reproduce this paper as a complete unit (no

extracts). All copies must include the copyright notice as follows:

Copyright © 1996 by Steve Lampa and Glenn Mazur All Rights Reserved.

Japan Business Consultants, Ltd. 1140 Morehead Ct. Ann Arbor, MI 48103 USA+1 (313) 995-0847 Fax: +1 (313) 995-3810 E Mail: [email protected]

Glenn Mazur

Updated tel fax email

Steve LampaBrand ExecutiveMarriott Hotels, Resorts, and Suites

Glenn H. MazurJapan Business Consultants, Ltd.

Abstract

Three recent trends have lead to changes in theway travelers view airport food; (1) healthier andlighter food, (2) more women travelers, and (3)fewer on-board meals being served. Host Marriott,which operates 70% of the U.S. airport food andbeverage market, wanted to assure that its productofferings were keeping up with customer demands.What they discovered was that their traditional ap-proach to new product and service developmentwas penny profit driven and not customer focused.QFD was employed to make quality and customersatisfaction more important. What ensued startledus all: within two weeks sales were up 50%, andafter one year sales had evened out at more thandouble their previous year’s level.

Key words: QFD, Service Quality, Food Prod-ucts, Bagels.

Company Profile

Host opened in 1897 as a purveyor of food, bever-age, news, and general merchandise in train sta-tions, the leading form of mass land transportationat that time. We have continued to serve that mar-ket by now controlling over 70% of the food andbeverage sales in U.S. airports. We also operatefood, beverage, and merchandise facilities in travelplazas on 12 east coast and midwestern highways.

Host currently commands $1.2 billion in sales peryear from its over 2,000 units in 170 locationsworldwide. Over 40 different types of regional andinternational branded products, such as BurgerKing, Taco Bell, TGI Fridays, etc., make up 65%of this business.

Our approach to developing new products andservices has been primarily localized, with eachoperation identifying the needs of its market,sourcing new products, testing them, and keepingthe ones that worked. That is, a loose, vaguely de-fined process. The two driving forces behind thiswere 1) to get a product that fit the category at thelowest price tag in order to drive the penny profitand cost of sales margins and 2) how much freeequipment the vendor would provide. Customerinput was not normally sought before or after that.Core items (coffee, hot dogs, baked goods, etc.)were secured through national contracts alsodriven by price and sales margins. Merchandisingand delivery to customers were handled in the tra-ditional way where we determined what was to bedone. Customer usage issues were not normallyconsidered.

In 1994, we began a strategic planning process toassess our strengths for the rest of the decade. Spe-cific competitive opportunities were identified thatexploited the competencies we had built up over

© 1996 Steve Lampa and Glenn Mazur All rights reserved.

1

Bagel Sales Double at Host MarriottUsing Quality Function Deployment

the last century. Several task forces were commis-sioned by Tom O’Hare, Vice President of Opera-tions, set up to explore ways to improve productquality. Some of the task forces focused on coreitems like hot dogs, baked goods, deli, etc. Wefelt, however, that given our widespread activitiesand the importance of these new business direc-tions, a more unified new product developmentprocess (NPD) was needed to assure that the qual-ity of the output could be maintained from thestrategic planning phase down through conceptand delivery of the service.

Through our work with GOAL/QPC, a Massachu-setts based quality training organization, we wereintroduced to quality function deployment and toGlenn Mazur of Japan Business Consultants andthe QFD Institute, one of the leading proponentsof the methodology in North America.

Why QFD for new serviceplanning and development?

QFD is designed to improve customer satisfactionwith the quality of our products and services.What can QFD do that is not already being doneby traditional quality systems? To understandQFD, it is helpful to contrast the differences be-tween modern and traditional quality systems.

Traditional Quality Systems

Traditional approaches to assuring quality oftenfocus on work standards [Love 1986], automationto eliminate people, or in more enlightened organi-zations, Quality Improvement Teams to empoweremployees to resolve problems.

As organizations are finding out, however, consis-tency and absence of problems are not enough of acompetitive advantage when the market shakes outsuboptimal players. For example, in the automo-bile industry, despite the celebrated narrowing ofthe “quality” (read that fit and finish) gap betweenU.S. and Japanese makers, Japanese cars still win

the top honors in the J.D. Powers Survey of NewCar Quality.

Modern Quality Systems

QFD is quite different from traditional quality sys-tems which aim at minimizing negative quality(such as poor service, broken product). With thosesystems, the best you can get is nothing wrong -which is not enough when all the players are capa-ble. In addition to eliminating poor service, wemust also maximize positive quality (such as con-venience, enjoyment). This creates value.

Quality Function Deployment (QFD) is the onlycomprehensive quality system aimed specificallyat satisfying the customer. It concentrates on maxi-mizing customer satisfaction (positive quality) -measured by metrics such as repeat business. QFDfocuses on delivering value by seeking out bothspoken and unspoken needs, translating these intoactions and designs, and communicating thisthroughout the organization. Further, QFD allowscustomers to prioritize their requirements, bench-mark us against our competitors, and then direct usto optimize those aspects of our organization thatwill bring the greatest competitive advantage.What business can afford to waste limited finan-cial, time and human resources on things custom-ers don’t want or where we are already the clearleader?

History of QFD

Quality Function Deployment began thirty yearsago in Japan as a quality system focused on deliv-ering products and services that satisfy customers.To efficiently deliver value to customers, it is nec-

Nothing Wrong

≠Everything Right


2

essary to listen to the “voice” of the customerthroughout the product or service developmentprocess. The late Dr. Shigeru Mizuno, Dr. YojiAkao, and other quality experts in Japan devel-oped the tools and techniques of QFD and organ-ized them into a comprehensive system to assurequality and customer satisfaction in new productsand services [Mizuno and Akao 1994, Akao 1990].

In 1983, a number of leading North Americanfirms discovered this powerful approach and havebeen using it with cross-functional teams and con-current engineering to improve their products, aswell as the design and development process itself[Akao 1983, Sullivan 1986, King, 1987]. Serviceorganizations have also found QFD helpful. Theauthor used QFD in 1985 to develop his Japanesetranslation business, Japan Business Consult-ants, and saw revenues increase 285% the firstyear, 150% the second year, and 215% the thirdyear [Mazur 1993c]. QFD was an important part ofFlorida Power & Light ’s successful bid to be-come the first non-Japanese Deming Prize recipi-ent in 1990 [“Quality System Implementation...”1988, Webb 1990, Bodziony 1995] and in the1994 Deming Prize awarded to AT&T PowerSystems. It has been successfully applied in theU.S. healthcare industry since 1991 at The Uni-versity of Michigan Medical Center [Gaucher1991, Ehrlich et al 1993, Ehrlich 1994], BaptistHealth System [Gibson 1994, 1995], and otherleading institutions. Interesting service applica-

tions also include the author’s development of anengineering TQM curriculum at The Universityof Michigan College of Engineering and the ap-plication to employee satisfaction and quality ofwork life at AGT Telus [Harries et al 1995]. Eachyear new applications are being reported in smallbusinesses as well [Mazur 1993c, 1994a]. Since1990, the author has consulted with other serviceorganizations in distribution, education, food serv-ice, personnel, finance, healthcare, repair, retail,and transportation businesses.

Early applications of QFD in service organizationsin Japan by Ohfuji, Noda, and Ogino in 1981 werefor a shopping mall, a sports complex, and a vari-ety retail store [Akao, 1990]. More recently,Kaneko has been integrating QFD, reliability, andquality circle activities in hotels, shopping centers,and hospitals [Kaneko 1990a, 1990b, 1991, 1992].

QFD has been heralded for such benefits as pro-moting cross-functional teams, improving internalcommunications between departments, and trans-lating the customer’s needs into the language ofthe organization.

Types of Requirements

To satisfy customers, we must understand howmeeting their requirements effects satisfaction.There are three types of customer requirements toconsider (see Figure 1) [Kano, et. al., 1984].

Revealed Requirements are typically what we getby just asking customers what they want. These re-quirements satisfy (or dissatisfy) in proportion totheir presence (or absence) in the product or serv-ice. Fast delivery would be a good example. Thefaster (or slower) the delivery, the more they like(or dislike) it.

Expected Requirements are often so basic thecustomer may fail to mention them - until we failto perform them. They are basic expectations with-out which the product or service may cease to beof value; their absence is very dissatisfying. Fur-ther, meeting these requirements often goes unno-ticed by most customers. For example, if coffee isserved hot, customers barely notice it. If it’s cold


3

Figure 1. The Kano Model (adapted).Service businesses must meet all three types of re-quirements - not just what the customer says.

Satisfaction

Dissat isfaction

RequirementFu lfilled

RequirementUnfulfilled

Expected (unspoken)

RevealedExciting (unspoken)

or too hot, dissatisfaction occurs. Expected re-quirements must be fulfilled.

Exciting Requirements are difficult to discover.They are beyond the customer’s expectations.Their absence doesn’t dissatisfy; their presenceexcites. For example, if caviar and champagnewere served on a flight from Detroit to Chicago,that would be exciting. If not, customers wouldhardly complain. These are the things that wowthe customers and bring them back. Since custom-ers are not apt to voice these requirements, it is theresponsibility of the organization to explore cus-tomer problems and opportunities to uncover suchunspoken items.

Kano’s model is also dynamic in that what excitesus today is expected tomorrow. That is, once intro-duced, the exciting feature will soon be imitatedby the competition and customers will come to ex-pect it from everybody. An example would be theability to have pizza delivered in thirty minutes.On the other hand, expected requirements can be-come exciting after a real or potential failure. Anexample might be when the passengers applaud af-ter a pilot safely lands the airplane in rough andstormy weather.

The Kano Model has an additional dimension re-garding which customer segments the target mar-ket includes. For example, the caviar andchampagne that’s exciting on the domestic flightmight be expected on the Concorde from NewYork to London. Knowing which customer seg-ments you serve is critical to understanding theirrequirements.

Thus, eliminating problems handles expected re-quirements. There is little satisfaction or competi-tive advantage when nothing goes wrong.Conversely, great value can be gained by discover-ing and delivering on exciting requirements aheadof the competition. QFD helps assure that ex-pected requirements don’t fall through the cracksand points out opportunities to build in excite-ment.

In summary, Kano found that the exciting needs,which are most tied to adding value, are invisible

to both the customer and the provider. Further,they change over time, technology, marketsegment, etc. The Japanese creators of QFDdeveloped tools such as the Voice of CustomerTables [Akao 1990b, Ohfuji et al 1990, Nakui1991, Marsh et al 1991, Mazur 1991a, 1991e,1992c, 1993a, 1993c] and coupled them to affinitydiagrams and hierarchy diagrams to break throughthis dilemna.

This process works best when the QFD team goesto gemba (where the customer interfaces with theservice) to observe, listen, and record the problemscustomers experience and the opportunities theywish to seize. The voice tables provided astructure for recording the data. Going to thegemba can be difficult for those who are used toseeing things from an internal point of view. Theytend to see more process problems and solutionsthan customer needs. Systematic tools can help theQFD team see the world from the customer’s pointof view.

Customizing the QFD roadmap for Host Marriott

Although Host had previously experimented withthe House of Quality, this was their first attempt atComprehensive Service QFD [Mazur 1993a]. Inorder to thoroughly examine all facets for applica-bility beyond the current project, they elected toexplore all the deployments in ComprehensiveService QFD. The standard deployments are ex-plained in Table 1.

Getting executive buy-in

The customization process began in March, 1995with a one-day QFD overview presented to execu-tives of both Marriott International and Host. Mar-riott attendees included Sam Bonfe, Director ofCatering Standards MHRS, Jim Burns, Brand Ex-ecutive, Jeff Brindle, Mary Scott, and Griff Lind-say from the New Business Team, and HelenaLight-Hadley, Director of TQM MHRS. FromHost came Jim Boragno, Sr. VP Products andStandards, Suzie Hill, Director of F&B Standards,


4

Dick Knockerbocker, Director of Procurement,Bob Stanton, General Manager and OTL - Wash-ington, D.C., Cindy Lynch, Food and BeverageStandards - St. Louis, and Ed Rudis, GeneralManager and OTL - Minneapolis. The meetingwas hosted by the author, Steve Lampa, then VicePresident of TQM and featured Glenn Mazur asthe subject matter expert.

The purpose of this meeting was to expose theseexecutives to the methodology so that they couldparticipate in determining whether QFD should beadopted as the standard new product and servicedesign process. At this meeting, it was determinedto do a pilot QFD project at the Phoenix Sky Har-bor International Airport around improving bakedgoods products.

As a service, there are fewer large capital invest-ments than in manufacturing companies, and it ispossible to experiment in a “living lab” and makemodifications relatively quickly. QFD should betailored to address these simpler business needs.

'HSOR\PHQW 3XUSRVH

&XVWRPHU'HSOR\PHQW

7R GHWHUPLQH ZKLFK FXVWRPHUV DQG JHP�EDV DUH FULWLFDO WR RXU VXFFHVV�

9RLFH RI&XVWRPHU'HSOR\PHQW

7R XQGHUVWDQG WKH WUXH QHHGV RI WKH FXV�WRPHU WKURXJK DQDO\VLV RI VSRNHQ DQGXQSRNHQ UHTXLUPHQWV DQG FRQWH[W RI XVH�

4XDOLW\'HSOR\PHQW

3ULRULWL]DWLRQ RI FXVWRPHU QHHGV DQGWUDQVODWLRQ RI WKHP LQWR VHUYLFH PHDV�XUHV�

)XQFWLRQ'HSOR\PHQW

7R LGHQWLI\ DQG SULRULWL]H UHTXLUHG DQGQHZ DFWLYLWLHV ZLWKRXW LGHQWLI\LQJ KRZ WKH\DUH WR EH SHUIRUPHG�

5HOLDELOLW\'HSOR\PHQW

7R LGHQWLI\ DQG SUHFOXGH IDLOSRLQWV IURP DSURFHVV�

1HZ 3URFHVV'HSOR\PHQW

7R FRQFHLYH DQG VHOHFW DOWHUQDWLYH ZD\VWR SHUIRUP WKH DERYH IXQWLRQV�

7DVN 'HSOR\�PHQW

'HWDLOHG EUHDNGRZQ RI WKH VHOHFWHG SURF�HVV WR LGHQWLI\ UHVSRQVLELOLW\ DQG SHUIRUP�DQFH UHTXLUHPHQWV�

6WDQGDUGL]D�WLRQ

-RE GHVFULSWLRQV DQG VWDQGDUG RSHUDWLQJSURFHGXUHV WR KROG WKH JDLQV�

7DEOH �� 7KH VWDQGDUG GHSOR\PHQWV RI &RPSUH�KHQVLYH 6HUYLFH 4)'�

Not everything in our core businesses need to benew and innovative products, so this limited trialhelped us to find an appropriate depth.

The Phoenix team was lead by Wayne Eddy(Multi-unit Manager Terminal 3), with Terry Ell(General Manager), Pat Banducci (Controller),Michael Galvin (Multi-unit Manager Terminal 4),Joe Campbell (Commissary Manager), and How-ard Rudin (Cluster Marketing Manager of Phoenixand San Diego).

We did not want to take on all baked goods for thefirst project, and so our first job in QFD was to de-termine what would be best suited for productchange. We began this three month journey inMay, 1995.

The bagel project details

Customer Deployment

Since QFD, like most TQM activities, tries to fo-cus resources on the most important areas, it wasuseful to determine the key customers we neededto satisfy. The logic here was that if we could meetor exceed the most important expectations of themost important customers, the rest would take care


5

3URMHFW *RDOV

3URMHFW

*RDOV

7\SH RI

EDNHG JRRG

3URMHFW

*RDOV

8QLW 7\SH

7\S

HRI

EDN

HGJRRG

&XVWRPHU

6HJPHQW

8QLW7\SH

Figure 2. CustomerDeployment Matrices.

of itself. The generic model of customer deploy-ment [Mazur 1993a] flows from identifying andprioritizing project success criteria to identifyingand prioritizing core competencies to identifyingand prioritizing customer segments. Since we werealready dealing with strategic competencies, Ma-zur helped us redefine the customer deployment tofit our situation. Figure 2 is a matrix flow chart ofthat process and figures 3 and 4 give a portion ofthe details. The purpose of these matrices was todetermine key customers of key unit types (termi-nal, unit area, etc.), that would sell the targetedbaked good, that would lead to the project beingdeemed successful by management. Once identi-fied, we would target our market research on thesecustomer segments first, thus conserving our re-search activities to the most fruitful segments.

Our first task was to clearly define how the projectwould be deemed successful by our management.First we brainstormed and then used an interrela-tionship digraph (details omitted) to understandthe drivers and “resultors” of these goals. Wefound that customer satisfaction drove many of theother goals and should be the primary focus of theproject. Increased sales, profit improvement, land-lord satisfaction, associate satisfaction and 15other goals were identified. With an affinity dia-gram, they were grouped under 5 headers: cus-tomer satisfaction, associate satisfaction, landlord(airport authority) satisfaction, profit, and won andretained sales contracts. Some goals were moreimportant than others, and so a prioritization ma-trix [Brassard 1989] was used to prioritize them.See Figure 3.

The next step was to augment our traditionalbaked goods with other potential varieties in orderto identify the kinds of baked goods that mightlead to customer satisfaction in an airport setting.Following Mazur’s “batman” process of brain-storm, affinity, tree, matrix, a hierarchy of possi-ble baked goods was created. This assured that wehad not overlooked any baked goods that couldhave made an especially exciting offering. SeeFigure 4 for a portion of the tree.

Breads

Croissants

Bagels

Muffins

Scones

Specialty

Cinnamon rolls

Donuts

Figure 4. Tree of types of baked goods


6

Fig. 3. Prioritization of project goals. CS AS LL PI WR RAW % OFSCORE TOTAL

CUSTOMER SATISFACTION (CS) 1 5 10 5 10 31.0 40.5%ASSOCIATE SATISFACTION (AS) 0.2 1 5 5 10 21.2 27.7%LANDLORD SATISFACTION (LL) 0.1 0.2 1 0.2 5 6.5 8.5%PROFIT IMPROVEMENT (PI) 0.2 0.2 5 1 10 16.4 21.4%WIN & RETAIN CONTRACTS (WR) 0.1 0.1 0.2 0.1 1 1.5 2.0%

TOTALS 1.60 6.50 21.20 11.30 36.00 76.60 100.0%

The baked goods tree was prioritized in a matrixfrom the project goals and their priority weights(details omitted). The analysis was to determinewhich baked goods would contribute most to thesuccess of the project goals. From this bagels wereselected.

The next step was to determine the type of retailunit we would sell these in. The batman techniquewas used to details these and a matrix was createdwith the highest priority baked goods and the typeof unit (details omitted). The unit types includedfull service restaurant, concourse kiosk with largedisplay cases, kiosks with small display cases, andbranded outlets. From this matrix we learned thetype of sales unit which would be most successfulat selling bagels - the concourse kiosk with largedisplay cases.

The next phase was to identify customer segmentsbased upon use characteristics such as time of day,purpose of coming to airport, etc. This was achange from the usual market research that deliv-ers demographic characteristics, such as income,education, etc. The batman process was applied toorganize these into a matrix (details omitted) withthe unit type based upon what type of customerwas most likely to eat at a concourse kiosk. Thehighest priority customer segment turned out to bewomen traveling in the morning on business. Wedecided to look at both men and women. This ishow we selected the gemba. The next step was togo to the gemba and determine the needs of thesekey customers.

In our traditional approach to going to the gembain the context of product planning, our attentionwas on internal issues such as sanitation, staffinglevels, product display, etc, rather than on the cus-tomer using our products and facilities. Mazurtook us on a practice run down to the cafeteria,where the team spent about an hour observing cus-tomers enter (or choose not to enter) the cafeteria,look around for menus, inspect the food, takethings, put things back, pay, and try to find a table.What we were taught to look for and record wereevidence that the customer was able to completeeach action easily and pleasurably and to identify

barriers. Especially valuable were smiles andgrimaces on the faces of our customers.

What we got was a view we don’t normally see inour planning sessions in meeting rooms. Later, atthe bagel gemba, there were many usage issues wehad not seen before. For example, the packagedcream cheese was difficult to open, plastic utensilsbroke, there was no place to sit. Careful analysisof this data and interviews revealed that more ba-gel varieties and flavored cream cheeses were de-sired. We also noticed that we were selling bagelsin a way that focused on speed of service (wewouldn’t cut bagels or toast them which wethough could hold up the line), so we didn’t offerthe most popular ways bagels are eaten! Our man-agers didn’t believe customers really wanted atoasted bagel because they never asked for them.Boy, were we wrong!

Voice of Customer Deployment

At the gemba, the spoken words and observed ac-tions of the customer were recorded in the Voiceof Customer Tables Part 1 and 2, which record us-age data, such as time of day, whether meal orsnack, etc. and sort the voices into benefits vs. fea-tures, respectively. The benefits, called demandedqualities in QFD, were put through the batmanprocess, and included items such as “I have morechoices,” “Tastes good,” “Easy to carry,” etc.

A survey was conducted of bagel eaters at thegemba and about 50 responses were received.Demographics were about 40% men to 60%women, about evenly split between Phoenix resi-dents and those who were not, and were abouttwice as many pleasure travelers as business trav-elers. They were asked to prioritize these benefitsso that we would know which they valued themost. The survey also asked them to compare thecurrent bagel offered at our airport terminals withthose they had elsewhere in terms of each of thebenefits. Frequency distributions of responseswere incorporated into an analysis that showed uswhat was most important to customers and wherecompetition was perceived as better. Our missionthen became to exceed the competition in those


7

benefits which were most important to the cus-tomer. A portion of the survey is shown in table 2.

7DEOH �� 6XUYH\ WRWDOV RI FXVWRPHUV· SUHIHUHQFHV DQG

FKRLFHV

Quality deployment

The next phase was to translate the customer bene-fits into service measurements and performancetargets that would help us design the new bagelservice right the first time. QFD uses a special ma-trix called a House of Quality for this purpose.The House of Quality brings together on one sheetof paper the customer needs, preferences, andchoice data from the market survey, the company’sresponse to those needs in terms of service meas-ures and performance targets, and yields a priori-

tized set of areas the company needs to addressfirst.

To save time, we approached the House of Qualityat two levels. First, we used just the general cate-gories of customer benefits or demanded qualities,and then we made a second House detailing onlythe most important areas from the first House. Fig-ure 5 shows the latter of these.

The top four performance measures were selectedfor improvement efforts; the rest were to be main-tained at the current level of performance. Thesefour were giving 50-60% of a display case to ba-gels, increasing the number of bagel varieties fromtwo to six, increasing the number of topping varie-ties from three to five, and adding a toaster optionto heat the bagels at time of serving.

Function Deployment

Once the performance targets were specified, itwas necessary to determine what activities wouldbe affected and who would be responsible formaintaining these performance levels. In Compre-hensive Service QFD, these are identified throughfunction deployment. Again, a two step approach,first a general matrix then a specific one was gen-erated to identify the business functions thatwould be necessary to implement the changes. The


8

)LJXUH �� 3KRHQL[ %DNHU\ 3URMHFW +RXVH RI 4XDOLW\

'(0$1'(' 48$/,7,(6 Y� 48$/,7< $775,%87(6 �*5$1'&+,/'5(1�

48$/,7< $775,%87(6

'(0$1'(' 48$/,7,(6 /,*+7,1*

63$&(�',0(16,21

9,6,%,/,7<2)

237,216

�2)%$*(/

9$5,(7,(6

6,=(

)/$925,17(16,7<

$520$

&2/25

6+$3(

�2)7233,1*

9$5,(7,(6

635($'$%,/,7<

3257,216,=(

�2)+($7,1*

237,216

�2)&877,1*

237,216

�2)6(59*

,03257$1&(

&855(17

3/$1

,03529(07

6$/(632,17

$%62/87(

'(0$1'('

, +$9( 025( &+2,&(6�

%$*(/6 � � � � � � � � � � � � � � � ��

7233,1*6 � � � � � � � � � � � � � � � ��

35(3� 237,216 � � � � � � � � � � � � � � � ��

6$7,6),(6 0< $33(7,7(�

7$67(6 *22' � � � � � � � � � � � � � � � ��

9,68$//< $33($/,1* � � � � � � � � � � � � � � � ��

3/($6$17 $520$ � � � � � � � � � � � � � � � ��

$%62/87( :(,*+7 ��

48$/,7< $775,%87( :(,*+7 � � ��

285 &855(17 3(5)250$1&( ��

7$5*(7 �� a� �a� �

81,7 2) 0($685(0(17 � FDVH � YDUL� � YDUL� �

VSDFH HWLHV HWLHV RSWLRQV

matrix deployed the quality attributes and theirpriority weights from the bottom of the House ofQuality to prioritize the most critical businessfunctions for assuring success of the new bagelservice. It showed that the commissary, sourcing,acquiring, and shipping of the new products wouldbe critical. At the operations level, displaying andmaintaining the attractive appearance of the bagelswas identified as critical. See Figure 6. Thesourcing people began right away contacting thebig bagel vendors and cream cheese producers tolearn what the current favorites were. Mazur intro-duced us to an industrial toaster company, PrinceCastle, that had a toaster that could toast the bagelin about the same time as it takes to complete thesale, so there would be no delay of the airline pas-senger.

Reliability deployment

When developing a new service, it is important toassure that any new processes employed are not

failure prone prior to the new service starting up.Comprehensive QFD uses reliability deploymentto first identify the potential fail points, and thento eliminate them in the design of the service. Fig-ure 7 is the reliability matrix. The failpoints areprioritized by deploying the demanded qualitiesfrom the House of Quality into the reliability ma-trix. The highest ranking failpoint in this matrix isrunning out of product.

New concept deployment

At this point in the QFD study, we now under-stood customer preferences and choices, perform-ance targets, key business functions, and potentialfailures to avoid. This gave us the information tobegin developing alternative processes to fulfillingthese requirements.

Different display cases, heating equipment,sourcing of bagels and cream cheese were workedinto process flows and examined for their ability


9

Figure 6. Phoenix Bakery Project Function DeploymentQUALITY ATTRIBUTES V. FUNCTIONS

QUALITY ATTRIBUTESFUNCTIONS

&200 !!! 236 !!!

6285&( $&48,5( 6725( 35(3$5( 6+,3 ),1,6+ 6(// 6(59,&( ',63/$< 0$,17$,1 &/($1 4�$�

352'8&7 352'8&7 352'8&7 352'8&7 352'8&7 35(3 352'8&7 &86720(5 352'8&7 352'8&7 (19,521� :(,*+76

9,6,%,/,7< 2) 237,216 � � � � � � � � � � � ��

�2) %$*(/ 9$5,(7,(6 � � � � � � � � � � � ��

�2) 7233,1* 9$5,(7,(6 � � � � � � � � � � � ��

�2) +($7,1* 237,216 � � � � � � � � � � � ��

$%62/87( :(,*+7 ��

4$�)81&7,21 :(,*+7 ��

)LJXUH �� 3KRHQL[ %DNHU\ 3URMHFW )XQFWLRQ 'HSOR\PHQW

'(0$1'(' 48$/,7,(6 Y� )$,/85( 02'(6 �&+,/'5(1�

)$,/85( 02'(6 !!!!!

'(0$1'(' 48$/,7,(6 %$*(/6 7233,1*6 :521* )$,/(' 728*+ 581 287 2) 581 287 2) 581 287 2) 0(66< (48,3017 581 287 2) ,*125(6 %$' +<*1(� '(0$1'('

%8517 )$// 2)) 3257,216 4$ 67'6� %$*(/ 6(59,&(:$5( &21',0(176 352'8&7 (19,52107 )$,/6 &/($1* 683/ &86720(56 '5(66 48$/,7< :7�

, +$9( 025( &+2,&(6�

%$*(/6 � � � � � � � � � � � � � ��

7233,1*6 � � � � � � � � � � � � � ��

35(3� 237,216 � � � � � � � � � � � � � ��

6$7,6),(6 0< $33(7,7(�

7$67(6 *22' � � � � � � � � � � � � � ��

9,68$//< $33($/,1* � � � � � � � � � � � � � ��

3/($6$17 $520$ � � � � � � � � � � � � � ��

$%62/87( :(,*+7 ��

)$,/32,17 :(,*+7 ��

to meet the requirements. As mentioned before,Mazur introduced us to Prince Castle, a companythat uses QFD to develop its commercial kitchenequipment and from their product line, the Excali-bur conveyor toaster was selected for its speed andsafety. Several bagel sourcing options were con-sidered including having bagels delivered by a lo-cal bakery twice a day, baking them at thecommissary and delivering them to the concoursekiosks on a pre-set or on a per request basis, andbaking them on site. New cutting devices werealso explored.

After analyzing the alternative concepts, a par-baked and frozen bagel supplied by Uptown Ba-gels that can be thawed quickly and baked in thekiosk in six minutes and would allow fresh bakedbagels produced to business demands was se-lected. A mandatory selection of plain, onion, cin-namon raisin, and honey wheat were chosen alongwith an optional blueberry and flavor of themonth. Different cream cheese options were alsoexplored. The selected process called for mixingflavors into a cream cheese that is whipped andblended on premise and then pre-packaged. Man-datory flavors are plain, onion/chive, and gardenvegetable with options of strawberry, low fat, anda flavor of the month. Appropriate signage wasalso developed.

Task deployment

The best laid plans come to fruition when indi-viduals are made responsible for carrying out spe-cific tasks in a manner that achieves the targets

that were designed and planned in the previoussteps. This is what task deployment does. Mazurdeveloped this deployment from the process qual-ity control sheets commonly found in manufactur-ing. The purpose of task deployment is to assignfor each step in the newly developed process, theperson responsible, timing, location, equipment,skill or training, performance, and self inspectionrequirements. When these are met, then the newprocess is assured that it can meet all the require-ments specified, and can thus meet the most im-portant customer benefits, which will lead toachieving the goals of the project as defined at thebeginning of the QFD. An excerpt of the task de-ployment tables is given in Table 3.

Standardization

No project is complete until we can assure the on-going performance of the new system. Since QFDis a Total Quality Management approach, stan-dards should be created for initial and ongoingtraining of employees and associates, vendor com-pliance, etc. After the tests were completed inAugust, the QFD team began to develop standardsthat have been compiled in a booklet entitled SalesDouble: QFD Bagel Project - A QFD Approach.Published by Host Marriott Services at Phoenix,Arizona, this shows the specifications, procedures,policies, equipment, and expected results of thisnew service. This standard has been adopted byour F&B standards department as the standard forall our generic (non-branded) units that sell bagels.Excerpts are given in Figure 8.


10

:KDW :KR :KHQ :KHUH +RZ +RZ PXFK :K\ 2WKHU

FXWWLQJ EDJHOV 0LNH *DOYLQ7�� DVVRFLDWH

E\ $XJ� ��

LQ KRXVH WHVW HTXLSPHQW XQWLO FRPIRUWDEOH�EXW QR OHVV WKDQ ��EDJHOV

VSHFV IRU FXW�WLQJ PHWKRG

IDLOXUH PRGHVVDIHW\VSHHG RI VHUY�LFH LVVXHV

RUGHU EDJHOFXWWHU

-RH &DPSEHOO E\ $XJ� �� LQ KRXVH SXUFKDVH RUGHU DW OHDVW � IRU HDFKWHVWLQJ XQLW SOXVRQH EDFNXS

IRU VWDUW RI SUR�MHFW LQ XQLW

IDLOXUH PRGHVSURSHU NQLIHNQLIH OHQJWK

VFRRSV IRU WRS�SLQJ SRUWLRQLQJ

-RH &DPSEHOO E\ $XJ� �� LQ KRXVH WHVW RUGHU � RI HDFK IRU HDFKWHVW XQLW

GLIIHUHQW W\SHVRI WRSSLQJV�SRUWLRQ FRQWURO�VSHHG RI VHUY�LFH

SRVVLEOHHTXLRSPHQWRWKHU WKDQVFRRS� EUHDN�DJH

7DEOH �� ([FHUSW IURP WDVN GHSOR\PHQW WDEOH IRU 3KRHQL[ EDNHU\ SURMHFW�

Sales results

As the title implies, sales in the Phoenix airportconcourse kiosks more than doubled as a result ofthe QFD efforts. Table 4 shows some of the re-sults.

/RFDWLRQ $YJ�

'DLO\

VDOHV EH�

IRUH 4)'

6DOHV ��

GD\V DI�

WHU 4)'

6DOHV ��

GD\V DI�

WHU 4)'

�

FKDQJH

7HUPLQDO � 1: ��

7HUPLQDO �

0DLQ 6QDFN %DU

��

7DEOH �� 6HOHFWHG UHVXOWV RI 3KRHQL[ EDJHO SURMHFW�

Changes in customer choice data were alsotracked. Figure 9 shows for some of the most criti-cal customer benefits how our product was per-ceived compared to other bagels customers havehad. As shown there were dramatic improvementsin the key customer benefits of more bagel andcream cheese variety and heating options, as wellas other benefits such as tastes good and easy tospread. And we were pleased that the enhancedoptions actually yielded improvements in servicespeed, which was contrary to what we initiallyworried would deteriorate.

Conclusions

The benefits of QFD were certainly proved in thisproject. That we were able to achieve two to threetimes sales growth in only one month and thensustain that over the next six months speaks to thestaying power of focusing on the customer andthen standardizing the resulting improvements.Through this careful analysis, job responsibilitieshave been improved and those who are responsible

for the day-to-day operations are involved in theprocess.

Our next step is to carry this to our San Franciscooperations, and to build this process into other fu-ture operations (see next section). The time com-mitment still concerns us in terms of a cost-benefitratio. By the end of the project, six people hadspent fifty-three hours each (a total of 318 hours)on this QFD project. Three issues emerge that af-fected this.

1. The team was new to QFD and manyof its tools. In the future, the 7 Man-agement and Planning tools should betaught first.

2. We wanted to explore the full powerof Comprehensive Service QFD in or-der to judge its applicability to futureprojects. Thus, some of the steps wetook were more for learning purposes.

3. We wanted to involve the corporateTQM department and so the projectwas stretched out over three months.

Now that we are educated in the process, we in-tend to review the process steps in order to de-velop a “down and dirty” approach that can bedone by a team dedicating their time off-line,rather than trying to fit it into their regular day.

Mazur has recommended Blitz QFD, a matrix-lessapproach developed by Mark McDonald of Ander-sen Consulting and Richard Zultner of ZULTNER& CO. for software QFD and Glenn Mazur forservice QFD. It is quick enough that it can be donein a single day, if the team members have assem-bled the correct data in advance.

In the end, our plan is to develop a standard ap-proach that must be followed in order to changethe product specifications of our core products.

Future activities: Halo Effect

The benefit of going to the gemba to understandthe customer’s perspective is being integrated into


11

our planning process at Phoenix Sky Harbor Inter-national Airport. Two projects have been engagedsince January. One was the design of a new sand-wich deli. Strong customer demands such as“sandwich is made the way I like it,” “sandwich isfresh” and “sandwich is fast,” have been realizedwith a moving sandwich line where the sandwichis moved from station to station by the attendantswho add ingredients chosen by the customer. Thishas reduced a line back up as well as increasedsatisfaction by being able to choose what they get.Also, to make it easier for both the attendants toload bottled drinks in the refrigerator and for cus-tomers to remove them, the sliding door refrigera-tor has been replaced by one with an air curtainsimilar to units in a supermarket. Drink salesjumped almost immediately, and this is now be-coming a standard configuration at other Host air-port properties in the Southwest.

Acknowledgments

QFD is a cross-functional process that enlists thehelp of many people. I would like to thank fore-most, Wayne Eddy and his team at the PhoenixSky Harbor International Airport for their dedi-cated work on this project. It was by them and forthem that this project succeeded. I would also liketo thank the other executives at Host Marriott whotook their time to study this process with me. Thebagel, cream cheese, and toaster vendors also de-serve acknowledgment for sharing their data andfor helping with testing of new configurations. Fi-nally, thanks to Glenn Mazur for his guidance andwillingness to customize the process on the fly asour needs became clearer through the process, andfor his off-site support of our progress via fax andphone.

References

Akao, Yoji, ed. 1991. Hoshin Kanri: Policy Deploy-ment for Successful TQM. Translated by Glenn Ma-zur. Cambridge, MA: Productivity Press. ISBN0-915299-57-7

Akao, Yoji, ed. 1990a. Quality Function Deployment:Integrating Customer Requirements into Product

Design. Translated by Glenn Mazur. Cambridge,MA: Productivity Press. ISBN 0-915299-41-0

Akao, Yoji, ed. 1990b. Introduction to Quality Deploy-ment. (In Japanese) Tokyo, Japan: JUSE. ISBN4-8171-0257-8

Akao, Yoji. 1983. Company-Wide QC and Quality De-ployment. Chicago, IL: The Cambridge Corpora-tion.

Baluja, Sanju, Brian Elliott, Eric Endsley, Jim Kaounas,Robert Lepler, Elizabeth Lewis. 1995. Shalimar,Haute Cuisine of India: Total Quality ManagementGuide. University of Michigan.

Bodziony, Bob. 1995. “QFD and Deming Prize Activi-ties at FPL.” Transactions from the Seventh Sympo-sium on Quality Function Deployment. Ann Arbor,MI:QFD Institute.

Brassard, Michael. 1989. The Memory Jogger Plus. Me-thuen, MA: GOAL/QPC. ISBN 1-879364-02-6

Brown, S.W., E. Gummesson, B. Edvardsson, and B.Gustavsson. 1991. Service Quality: Multidiscipli-nary and Multinational Perspectives. Lexington,MA: Lexington Books. ISBN 0-669-21152-4

Ehrlich, Deborah. 1994. “Health Care: Tailoring a Serv-ice Industry.” Transactions from the Fourth Sympo-sium on Quality Function Deployment. Ann Arbor,MI:QFD Institute.

George, William R. and Barbara E. Gibson. 1991.“Blueprinting: A Tool for Managing Quality inService.” Brown, S.W., E. Gummesson, B. Ed-vardsson, and B. Gustavsson. 1991. Service Qual-ity: Multidisciplinary and MultinationalPerspectives. Lexington, MA: Lexington Books.pp. 73-91. ISBN 0-669-21152-4

Gibson, Jeff. 1995. “Happy Feet Part II: Return of thePrinceton Foot Clinic.” Transactions from theSeventh Symposium on Quality Function Deploy-ment. Ann Arbor, MI:QFD Institute.

Gibson, Jeff. 1994. “Health Care Services: PrincetonFoot Clinic.” Transactions from the Fourth Sympo-sium on Quality Function Deployment. Ann Arbor,MI:QFD Institute.

Harries, Bruce and Matthew Baerveldt. 1995. “QFD forQuality of Work Life.” Transactions from the Sev-enth Symposium on Quality Function Deployment.Ann Arbor, MI:QFD Institute.

Kaneko, Noriharu. 1992. “QFD Applied in a Restaurantand Hotel Chain.” Transactions of the 1st Euro-


12

pean Conference on Quality Function Deployment.Milan, Italy.

Kaneko, Noriharu. 1991. (In Japanese) “Using QFD in aService Industry.” Transactions of the First (Japa-nese) Symposium on Quality Function Deployment.Tokyo, Japan: JUSE.

Kaneko, Noriharu. 1990a “Quality Assurance in ServiceIndustries.” ASQC Congress Transactions. SanFrancisco, CA: American Society for Quality Con-trol.

Kaneko, Noriharu. 1990b. “QFD at a Supermarket.”Case Studies. Austin, TX: Kaizen Institute ofAmerica 3-day Public Seminar.

Kano, Noriaki, Nobuhiko Seraku, Fumio Takahashi, andShinichi Tsuji. 1984. “Attractive Quality and Must-Be Quality.” Translated by Glenn Mazur. Hinshitsu14, no. 2. (February): 39-48. Tokyo: Japan Societyfor Quality Control.

Kawakita, Jiro. 1986. (In Japanese) The KJ Method:Seeking Order Out of Chaos. Tokyo: Chuokoron-sha. ISBN 4-12-001517-3

King, Bob. 1987, 1989. Better Designs in Half theTime: Implementing QFD Quality Function De-ployment in America. Methuen, MA: GOAL/QPC.

Love, John F. 1986. McDonald’s: Behind the Arches.Toronto: Bantam Books. ISBN 0-553-05127-X

Marsh, S., J.W. Moran, S. Nakui, and G. Hoffherr.1991. Facilitating and Training in Quality functionDeployment. Methuen, MA: GOAL/QPC. ISBN1-879364-18-2

Mazur, Glenn. 1994a. “QFD for Small Business: AShortcut through the ‘Maze of Matrices’,” Transac-tions from the Sixth Symposium on Quality Func-tion Deployment, Ann Arbor, MI:QFD Institute,June, 1993.

Mazur, Glenn. 1993a. Comprehensive Service QualityFunction Deployment. Ann Arbor, MI: Japan Busi-ness Consultants, Ltd.

Mazur, Glenn. 1993c. “QFD for Service Industries:From Voice of Customer to Task Deployment,”Transactions from the Fifth Symposium on QualityFunction Deployment, Ann Arbor, MI:QFD Insti-tute, June, 1993.

Mazur, Glenn. 1992a. “Task Deployment for ServiceQFD and the Americans With Disabilities Act,”

Proceedings from the GOAL/QPC Ninth AnnualConference, Boston, MA.

Mazur, Glenn. 1992b. “The Seven Management andPlanning Tools and QFD,” Annual Review of Com-munications, National Engineering Consortium,Vol. 46, 1992-93, pp. 804-811.

Mazur, Glenn. 1992c. “Voice of the Customer Table,”Transactions from the Fourth Symposium on Qual-ity Function Deployment, Ann Arbor, MI:QFD In-stitute.

Mazur, Glenn. 1992d. Basics of Quality Function De-ployment. Ann Arbor, MI: Japan Business Consult-ants, Ltd.

Mazur, Glenn. 1991a. “Voice of Customer Analysis,”Proceedings from the GOAL/QPC Eighth AnnualConference, Boston.

Mazur, Glenn. 1991b. “Sensory QFD: Matching Prod-uct Image to Customer Image Using Kansei Engi-neering,” Proceedings from the GOAL/QPC EightAnnual Conference, Boston, MA.

Mazur, Glenn. 1991c. “Quality Function Deployment:Should You Really Function Without It?” TheQuality Management Forum, ASQC, Vol. 17, no.3.

Mazur, Glenn. 1991d. “QFD - The House of Quality:Practicing What We Preach,” OMNEX News, AnnArbor, MI, Fall, 1991.

Mazur, Glenn. 1991d. “Getting Started with QualityFunction Deployment,” OMNEX News, Ann Arbor,MI, Summer, 1991.

Mazur, Glenn. 1991e. “Voice of the Customer Analysisand Other Recent QFD Technology,” Transactionsfrom the Third Symposium on Quality Function De-ployment, Ann Arbor, MI:QFD Institute, June,1991.

Mizuno, Shigeru and Yoji Akao, ed. 1994. QualityFunction Deployment: The Customer-Driven Ap-proach to Quality Planning and Deployment.Translated by Glenn Mazur. Tokyo:Asian Produc-tivity Organization. ISBN 92-833-1122-1

Mizuno, Shigeru, ed. 1988. Management for QualityImprovement: The 7 New QC Tools. Cambridge,MA: Productivity Press. ISBN 0-915299-29-1

Nagamachi, Mitsuo. 1993. The Art of Kansei Products.(in Japanese) Tokyo:Kaibundo. ISBN 4-303-72820-9


13

Nagamachi, Mitsuo. 1992. “Kansei Engineering and itsMethods.” (in Japanese) Keiei Systems. Vol. 2 No.2.

Nagamachi, Mitsuo. 1989a. “QC Activities for ServiceIndustries: Kansei Quality Control.” (in Japanese)Engineers, 11/89 Tokyo.

Nagamachi, Mitsuo. 1989b. Kansei Engineering. (inJapanese) Tokyo:Kaibundo. ISBN 4-303-71320-1

Nakui, Satoshi. 1991. “Comprehensive QFD System.”Transactions from the Third Symposium on QualityFunction Deployment. Ann Arbor, MI:QFD Insti-tute.

Nakui, Satoshi and Yoji Akao. 1990a. “DeterminingDemanded Quality Through Questionnaires.” Tone,K. and R. Manabe, ed. (In Japanese) Case Studiesin AHP. Tokyo: JUSE Press. pp. 136-145. ISBN4-8171-5016-5

Nakui, Satoshi and Yoji Akao. 1990b. “Function De-ployment Applied to Manufactured Housing.”Tone, K. and R. Manabe, ed. (In Japanese) CaseStudies in AHP. Tokyo: JUSE Press. pp. 145-159.ISBN 4-8171-5016-5

Ohfuji, Tadashi, Michiteru Ono, and Yoji Akao. 1994(in Japanese) Quality Deployment Methods (2):Comprehnsive Deployment for New Technology,Relaibility, and Cost. Tokyo: Union of JapaneseScientists and Engineers. ISBN 4-8171-0259-4

Ohfuji, Tadashi, Michiteru Ono, and Yoji Akao. 1990.(in Japanese) Quality Deployment Methods: Proce-dures and Practices for Making Quality Charts.Tokyo: Union of Japanese Scientists and Engineers.ISBN 4-8171-0258-6

Porter, Michael E. 1985. Competitive Advantage: Cre-ating and Sustaining Superior Performance. NewYork: The Free Press. ISBN 0-02-925090-0

“Quality System Implementation at Florida Power &Light Company.” Presented to Professor YojiAkao, March 28, 1988.

Saaty, Thomas L. 1990. Decision Making for Leaders:The Analytic Hierarchy Process for Decisions in aComplex World. rev. 2nd ed. Pittsburg:RWS Publi-cations. ISBN 0-9620317-0-4

Takasu, Hisashi. 1994. “QFD for Demanded Services:A Restaurant Case” Standardization and QualityControl. (in Japanese) Tokyo: Japan Standards As-sociation. Vol. 47. Sept. 94. pp. 52-61.

Webb, Joseph L. and W. C. Hayes. 1990. “QualityFunction Deployment at FPL.” Transactions fromthe Second Symposium on Quality Function De-ployment, Ann Arbor, MI:QFD Institute.

Zemke, Ron and Dick Schaaf. 1989. The Service Edge:101 Companies That Profit from Customer Care.New York: Plume Books. ISBN 0-452-26493-6

Zultner, Richard E. 1993. “The AHP in QFD: PrioritiesThat Fit Your Project.” Transactions from theFifth Symposium on Quality Function Deployment,Ann Arbor, MI:QFD Institute, June, 1993.

About the authors

Steve Lampa is currently Brand Executive withMarriott Hotels, Resorts, and Suites. From 1991 to1996 he was Vice President of Total Quality Man-agement at Host Marriott, and prior to that spent16 years in operationsa and human resources atMarriott Hotels. Lampa earned his Bachelor’s De-gree in English at Northern Illinois University. Heis a Malcolm Baldrige National Quality AwardExaminer from 1995-1996.

Glenn H. Mazur has been active in QFD since itsinception in North America, and has worked ex-tensively with the founders of QFD on their teach-ing and consulting visits from Japan. His primaryfocus is in the service industry, as a manager forover 15 years in automobile repair and parts ware-housing, as a teacher, and as an owner of a trans-lating and consulting business he started in 1982.He is one of North America’s leaders in the appli-cation of QFD to service industries, sits on severaladvanced QFD research committees, and sits onthe steering committee of the Symposium on Qual-ity Function Deployment held annually in Detroit.He is also Executive Director of the non-profitQFD Institute and an Adjunct Lecturer of TotalQuality Management at the University of Michi-gan College of Engineering. He lectures and trainsin QFD worldwide.

Mazur holds a Master’s Degree in Business Ad-ministration and a Bachelor’s Degree in JapaneseLanguage and Literature, both from the Univer-sity of Michigan.


14

bagel sales double at host marriott using qfd · bagel sales double at host marriott ......

Documents