Innovation,OrganizationandTechnology1InnovationintheUSServiceSectorMi,-hae!P.Gallaher,AlbertN.Link,a

Routledge Studies in Innovation, Organization and Technology

1

Innovation in the US Service Sector Mi,-hae! P. Gallaher, Albert N. Link, a"dJeffrey E. Petrosa

Innovation in the US Service Sector

Michael P. Gallaher, Albert N. Link, and Jeffrey E. Petrusa

n

~~

Routledge Taylor&FranclsGroup

LONDON AND NEWYORI<'

Contents

First published 2006

by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN

List offigures List of tables Acknowledgments

Simultaneously published in the USA and Canada

by Routledge 270 Madison Ave, New York, NY 10016

I 'I

II IIu 1,\

I

,,~Routledge is an

imprin, sfthe Taylor & Francis Group,

, ~ 路~~~,?~ltiforma:lmsmeJs >:ii\::漏/2006:MichaelP. Gallaher, Alben N. Link, and Jeffrey E. Perruse

~~"Typese[ in Gataroond by ji'":Ncwgen Imaging Systems (P) Led, Chennai, India Printed and bound in Grear Britain by Biddles Ltd, King's Lynn All rights reserved. No part of this book may be reprinted or reproduced or utilised in any Iorrn or by any electronic, mechanical, or ocher means, [lOW known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers.

British LibraryCatalog"irtg in Publication Data A catalogue record for this book is available from the British Library

VI VII

ix

1

Introduction

1

2

Innovation in the service sector

6

3 Telecommunications industry

36

4

Financial services industry

51

5

Systems integration services industry

69

6

Research, development, and testing service industry

81

7

Dimensions of innovation and productivity growth

97

8

Public policies to enhance innovation

111

Notes Bibliography Index

120 124 131

Library of Congress CaJaiogi"g hi Publication Data A catalog record for this book has been requested

ISBl'10:0-415-39068-0(hbk) ISBmo: 0-203-96631-7 (ebk) ISBm3: 978-0-415-39068-2 (hbk) ISBl'13: 978-0-20.1-96631-0 (ebk)

i

.'

Today that view has changed. Services are a large part of the economy and

2

Innovation in the service sector

<)"rmation technologies have. in many service industries, revolutionized the

., y.business is conducted. However, use of the manufacturing R&D model "-;~ineasure service-sector innovation activities is not completely appropriate "is likely ro lead to biased estimates of innovative investments by the ice sector. ";B~cause the service-sector innovation process is a relatively new phenomenon

I Introduction Much of the service sector's growth since the early 1980s has been based on new services with significant knowledge conrent. A large share of the knowledge content in services is built on advances in hardware and software that were imported or purchased ÂŁTom the manufacturing sector. However, the service secror adds _, val,;~ in a dia:~rem waJi than the manufacruring sector-by integrating .}'ll!91~ed.,physical technology into sysrems. To add this value, service-sector J:~~:,~~a~e':ITiiideconsiderable' investments in capabilities such as systems-level i,o[eg~tjon. The impact of these investments has been substantial. R&D investments in service-sector information technology (In have generated an estimated rate of return of 196 percent, while noninformarion technology investments in

general have only created an 11 percem return (TASC, 1998).

For the US economy to continue to experience hisrorical rates of productivity growth, the future performance of the service sector will be critical.

Productivity growth is associated with applying resources to inventive and innovative initiatives as measured

by spending on R&D. These expenditures,

which currently represent only a few percentage points ofGDP, support both basic research and app.lied research, which includes research into new specialized products for sale to industries and into the development of processes and process improvements internal to the innovating organization.

Traditionally, analysis of productivity growth in the service sector and the development of more accurate R&D estimates for that sector have suffered relative to other senors of the economy. This is because economic output and producrivi ty measures were originall y developed in an era when services were a smaller share of the economy and the absence of more accurate information was not critical to policy making. In addition, because improvements to service outputs tend to be related to changes in quality rather than in

quantity, productivity improvements are very difficult to measure.

do.!.ess structured compared to the manufacturing sector's process, the nature ~¼¡~gnitude of service-sector innovative investments are less understood. . rent taxonomies, objectives, and research processes create confusion over ification and data collection. This confusion makes NSF's job of deterrnini;,;,how to quantify total investments in service-sector R&D and describing i'ovation outputs difficult. The lack of an adequate R&D taxonomy and a ensus framework for analyzing service-sector R&D and innovative activity pefS the development and evaluation of public policy.

onomies of service-sector innovations ovation and technological change in the service sector are increasingly

sendenr on internal service-sector R&D, in addition to technology iiired from the manufacturing sector (Pilat, 2001). However, there is ';~rn that current NSF R&D statistics do not fully capture the level of ~~vative activity 'being performed within the service sector or, indirectly, rate of change of innovative activity. NSF's indusrrial R&D survey reports ...:... manufacturing performed 62 percent and nnnmanufacturing performed ercenr of total indusrrial R&D in 2001. This distribution could be interfed to imply that manufacturing may be performing a disproportionate ;e of national R&D relative to the sector's contribution to economic "wth and may be doing a preponderance of the innovative activity in the nomy.

Recent studies have found that innovations in high-technology equipment ~:ve been increasingly the product of end-user R&D activities and less from '~ridor R&D. These end users, which are frequently service-sector firms, have

ltsthand nnderstanding of what the technology needs ate within an industry how to innovate the existing technology to ultimately improve Mmpetitive advantage (von Hippel, 1988). For example, systems integration c:~'~: an integral component of most service sectors' innovative activities. As '>~liown in Table 2.1, this process involves rusromizing components for specific

aria

"applications. The service-sector firms (or specialized consultants) are essen:ial in the integration process because of their detailed knowledge of the specific

applications.

?

Table 2.1 Systems inregrarion in selected US service firms Name

Application dmTiption

Aetna

• Insurance classification plan loss control system • Expert system for providing interactive assistance in solving problems such as health care management • Th~ee-level distributed control for networking input/output devices

Amazon.com

Ciricorp

Merrill Lynch

Sabre Group

• System and method for providing multimedia bookmarks for hypertext markup language • Secure method for communicating credit card data when placing an order on a nonsecure network • Secure method and system for communicating a list of credit card numbers over a nonsecure network • System and method for delivering financial services Distributed network agents • Check alteration detection system and method Integrated system for controlling master account and nested subaccount(s) • Securities trading workstation • Information aggregation and synrhesizarion system • System to predict optimum computer platform

Source: Compiled from informacion contained in the US Parenr and Trademark Office's Web patent database. Available online at http://www.uspro.gov. Taken directly from CSTB (2000, p. 84).

The term "innovation" has historically been used to encompass a wide range of processes that include both R&D- and non-R&D-related activities.

In a broad sense, innovation may be new products, new processes. or new organizational methods that are novel and add value to economic activity.

These developmems or changes are shaped by interactions between a firm and various other organizations, including suppliers, collaborators, competitors, customers, technological infrastructures, and professional networks and environments. A firm's innovation pattern depends on changes in the behavior of these organizations and the expectations of other organizations' behaviors (Gallouj and Weinstein, 1997; Hauknes, 1998).

In general, innovation has also been described as any change in the characteristics of new products in terms of service. competence. and/or technical knowledge, brought abour by evolution, variation, disappearance, appearance, association, or disassociation (Gallouj and Weinstein, 1997). In this light. innovation captures producr and service modifications that mayor may not be derived from R&D. To date. most researchers believe that an accurate model of service irmovarion is still absent from the literature (Howells, 2000b). With a few

exceptions, such as. for example, Barras's reverse product cycle (Barras. 1986). the extant literature related to service-sector innovation focuses on differentiating service activities from manufacturing innovation paradigms, as opposed to building on the specific nature of service-sector products and process.

This state of the literatures suggests a need to integrate the unique traits of service-sector innovation into the existing taxonomies and innovation paradigms. The goal is to capture more diverse types of innovation in conceptual models of innovation. including what was once two distinct segments of the economy but has over time become increasingly less disparate (Gallouj and Weinstein, 1997; Arnable and Palornbarini, 1998). Although existing taxonomies have begun to address how innovation occurs in service firms, much more work remains to be done in terms of modeling an innovation system that incorporates services. The model we posit in this chapter is only one step toward that end.

Recent efforts to incorporate the service secror into models of innovation include. for example. Pavitt's (984) taxonomies for classifying sectoral patterns of technological change. By dividing a national economy into three sectors-supplier-dominated, production-intensive, and science-based, Pavitt outlined a dynamic relationship between technology and service industries. Professional. financial, and commercial services are captured in the supplier-dominated category. However, the firms associated with this sector were primarily described as firms that expend few resources developing processes and products, usually having weak in-house R&D capabilities where most innovations come from the supplier of equipment and materials;' Soete and Miozzo (989) have taken the Pavitt model a step further by expanding on the supplier-dominated sector, offering two new classificarions: a production-oriented category and an innovative-specialty category. The production category includes those service firms performing large-scale processing and administrative activities and developing physical Or information networks. The specialized technology suppliers' category includes firms performing science-based activities to develop proprietary technology through innovation.

In addition to understanding the distinctions between manufacturing and service-sector innovative activities, a second line of research has focused on the differences between products (e.g. new services) and process (e.g. new organizational and delivery processes) within the service industry. Gallouj and Weinstein (997) make this distinction by dividing innovation for the service firm into two classifications: technical characteristics (front-office tangible technologies for the parr of a firm in direct contact with clients) and

")

~

II 1 1

process characrerisrics (both rangible and inrangible back-office technologies such as methods, working tools, and organizarional theory). Evidence from Sirilli and Evangelista (998) suggests that service-sector firms are able to distinguish between service and process innovations. These authors identify engineering) technical consulrancy, and computing services as the most innovative service sectors.

Sundbo (1997) offers a strategic innovation paradigm where the firm's strategy is the core innovation determinant. He offers empirical evidence through a study of Danish service firms where he breaks services into three categories: top strategic organizations, characterized by large to medium service-secror firms that are mass producing services; network organizations, described as a loosely tied association of small firms that innovate little on their own; and professional organizations) defined as collective action groups with shared interest in technology interoperability.

Evidence of the service sector's technological maturity has been established throughout the literature (Amable and Palornbarini, 1998; Pilat, 2001).

Service-sector firms are taking on more proactive roles in innovation activities and in some cases are leading the innovation process. Given the distributive and "dynamic nature of the innovation process, both manufacturing and service-sector-firms are. beginning to collaborate through bilateral and multi.iarerahnetworks. This. coJlaboration of firms is referred to as a distributed innovationprocess (DIP) and suggests that service-sector firms are beginning to take more of a leading role in innovation.

The largest contribution to growth in the service sector is from a small subset of all services known as knowledge-intensive business services (KlBSs) (e.g. telecommunications, IT, networking, and organizational consultancy).

In these areas, services are playing an increasingly important role in technological change and productiviry growth by promoting standards and systems integration. Researchers predict that the developed economies of the world will soon enter a postindustrial period in which services drive economic growrh (OECD, 2001a).

As summary, over the past two decades, the literature has attempted co develop a conceptual framework aimed at understanding how service firms innovate. In recent years, the discussion has turned from the sizeable differences between the two sectors to their convergence. However, a growing number of authors point to such firms as IBM andlor Siemens (both large mass-production service firms) as examples of traditional manufacturers who now have a dominant share of their business activities associated with the sale of services (Howells, 2000b; OECD, 2001b).

Over time the once sizeable distinction between manufacturing industries and service-producing industries is narrowing (Am able and Palornbarini, 1998;

/<:>

'.:86:1); 2001a), The literature suggests thar disringuishing between the , :... ~,:(agriculture and raw material extraction), secondary (manufactur'i'~~and:tertiary (services) sectors of the economy is obsolete, as more firms in, manufacturing and services adopt the practice of encapsulating phys. roducrs with services (Howells, 2000b). Xerox was one of the first maners. to bundle its products with full-scale service agreements. Today, ,,"service leasing directly from manufacturers is common, and the trend is , cred to grow in the future. ;. 'is'" broadening of scope is referred [Q as servicisation, or the trend in '", acruring to encapsulate the physical product in a shell of services - 'nance, monitoring, maintenance, and repurchase). Servicisarion is also . Wing trend in the automobile and aerospace industries, and, as it es- more pervasive and better understood, researchers will begin to nk how they characterize the service sector and its innovation activities.

~urement of service-sector innovation ,. -ht

empirical evidence demonstrates that service-producing industries

;~bVate and have been doing so increasingly over time (Sirilli and Evangelista, .~8禄 As national institutions began measuring innovation activities in the

'-ice'sector, the original models and data collection instruments were based 路C., understanding of the technology innovation process as it applied to ufacturing firms. However, because of the intangible nature of the service '-'or's output, measuring the productivity of R&D performed using th~ ~torical measures of innovation, such as new products or patents (Gallouj It"Weinstein, 1997), is difficult. The manufacturing sector adds value to : ts-R&D is an input in. the innovation process-by continuously ';.;: roving the materials and design of their products, whereas the service or applies accumulared knowledge ro build organizational models or srems a mote abstract output than in manufacturing (Iankowski, 2001). In resabsence of such appropriate merrics, any resulting measure of innovative .:' iviry would by definition be limited. ervice innovations draw less directly on scientific breakthroughs and are ten small or incremental in nature, this means that small changes can lead ~'hew applications or reorganization of an existing technology. or syst~m -alat, 2001). Some large service-sector firms actually hav.e an 1O~OV~t10n epartment that promotes and collects ideas. Howev~r, this or~anlZa[I~nal esign may be the exception rather than the rule. ServlCe-se.ctor rnnovanons e rypically based on both market-wide and consumer-speCIfic needs, sriInnovation in service-sector firms is generally not a sysremarrc process:md ften consists of spontaneous ideas developed internal1y to meet the real-timer,"

needs of a specific client. In contrast, innovation in manufacturing firms is typically highly structured, with a systematic approach ro developing new products following the product life cycle. Although there is an attempt to systematically organize and account for innovation across all sectors, measuring innovation in the service sector is extremely subjective because of the intangible narure of its products (Pilat, 2001). Measuring innovation is further complicated in the service sector because it OCCurs throughout the organizational process.

Patenting in the service sector is at times more difficult because of high visibility or the inexcludability of the product or process. If a manufacturing firm develops an innovative process or product, it can keep the process or product a secret by not allowing anyone outside the firm to see it. Servicesector firms offering intangible products have much higher visibility, which makes it hard to contain trade secrets and easier for other firms to imitate the product or process.

J,t~ J_

in. the service sector

.,c;. "

R&D~;is>"ah:'_,integral component of the innovation process. R&D is di'stingtrishedfromthe braader categoty of innovative activities in that R&D 路i"defiii&kteY.include.activities that systematically use research findings and expand-the frontier of knowledge. Based on this definition, many acrivities frequently included as innovation are excluded from R&D, such as market research and technology adoption/or imitation.

The classical definitions of R&D are grounded in creating an artifact or physical producr (Tether et al., 2001) and, therefore, are harder ro apply ro intangible outputs of services, such as methods or organizational theory. For example, Sundbo and Gallouji (1999) identify four major categories of service innovation-product, process, organizational, and market-not all of which are considered R&D. However, the difficult distinction in services between product and process, also referred to as coterminaliry, often makes it difficult ro interpret what is R&D and whar is non-R&D (Sirilli and Evangelista, 1998; Evangelista, 1999).

Innovation surveys have found that R&D accounrs for a much smaller share of activity and expertise as related to service-sector innovation compared to innovation in the manufacturing sector. For example, R&D typically accounts for about half of manufacturing innovation expenditures, whereas, on average, R&D accounts for only about one-fourth of service-sector innovation expenditures. The exceptions are telecommunications, computer services, and engineering services where R&D accounts for over three-fourths of expenditures related to innovative activity.

novation inputs (elements) that are not classified as R&D include market ch, training in innovation, adoption and adaptation of new technology, ~up;,activities, organizational changes, and incremental impacts. Such icharacrerize much related activity in service-sector firms. ,::,~useservice activities are generally labor intensive, investment in " an' capital costs often represents a larger share of total innovation ndirures. For example, sraff training for evolving computer systems or roducr offerings is a continual process in service industries. tomization is also a gray area for distinguishing. between R&D and &Ddnnovation activities. Service-producing industries commonly take hers ..developed in the manufacruring sector and add value to them by _ ling customized systems or networks. Frequently, a system is specifiE;:'.tailored to an individual client and irs assembly represents a unique r tid. However, it may be unclear if this is the development of a new and roved product, hence to be included as R&D, or the reapplication of '~j'ting:, -rnethods and processes in providing a service. Possibly new ledge and refined techniques are developed as part of most custom ms integration services that could be classified as R&D. quisition and integration of technology may also be an important ponent of innovation that is frequently not included as R&D. These iries are of particular importance to service-sector firms because much of &D, associated wit~ services is embodied in products acquired from :de, of the service sector. The issue becomes what share of acquisition and ration activities is R&D and what share is simply technology adoption . itation that is not classified as R&D. usiness interactions and joint product development between the service,- ucing industries and manufacturing industries represent an increasing d. -in product aod service innovation. For example, American Airlines )<.,played a significant role in developing the design specifications for . gs 777 series. However, much of AA's activities were likely conducted staff in market research divisions, and it is unclear what share of this rk)was or should have been counted as R&D expenditures. Similarly, Ieeommunicarions and financial service providers are integrally involved in ecifying components that go into their systems. eNSF has traditionally measured R&D activity in the nonmanufactur.g;,sector from responses to its annual survey of private businesses using its orm RD-l. Form RD-l is an instrument that was originally developed for easuring R&D activity in manufacturing firms (Link, 1996); thus, by its esign its applicability to service-sector firms is reasonably questionable. ;:For reporting purposes, R&D is defined slightly differently across different .:US and international agencies. However, most US agencies and many foreign

Ii

r

i 1

1 I'I

1 !'.~ .1.'

,!.,

~~;

I, I I",

'I"

,II,

I'1.,: ,I'

1.

agencies follow NSF's definition of R&D. Research is. defined as the systematic study directed toward fuller knowledge or, understanding of the subject studied, Development is defined as the systematic use of knowledge or understanding gained from research, directed toward the production of useful materials, devices, systems, or methods, including the design or development of prototypes or processes.

Research is further classified as either basic or applied, depending on the objectives of the investigator. Basic research is research directed coward increases in the knowledge or understanding of fundamental aspects of phenomena and of observable facts without specific application toward processes or products. This type of research is rypically limited to the federal, university, and nonprofit sectors. However, NSF does include basic research as a reporting option on Form RD-l. Expenditures for basic research include the cost of projects that adhere to the aforementioned definition, but the projects could originate in fields that are of potential business interest in the future. Applied research is research directed toward gaining knowledge that will meet a specific need. This includes research for specific commercial objectives. '路路'".Based'ori,NSF's definition, an activity is considered R&D if it is related to one' or.tmore -of the 'following goals: pursue a planned search for new '路路knowledge~;whetheror not the search has reference to a specific application; apply existing knowledge to problems involved in creating a new product or process,' including work required to evaluate possible uses; andlor apply existing knowledge to problems involved in improving a present product or process.

Inflation-adjusted R&D performed in industry over the years 1986 through 2001 in total and for nonmanufacturing (defined as total less manufacturing R&D) is illustrated in Figure 2.1 2 In 2001, the nonmanufacturing share of total industry R&D was nearly 40 percent. That percentage had increased steadily over time. Nonrnanufacruring R&D as a percentage of total R&D has increased from about 8 percent in 1986 to irs current level. This NSF-reported dramatic increase in reported nonmanufacturing R&D is partly due to improved sampling procedures to collect nonmaoufacturing sector data, and to increased innovative behavior in nonmanufacturing firms.

The NSF has collected and published industrial R&D statistics since 1953.

Historically, a sample of firms was selected every 4 to 6 years to complete Form RD-l. In the intervening years, a subsample of only the largest firms was surveyed. In the early years of the survey, R&D was performed only in a small number of industries, and R&D in those industries was reported along wirh a catch-all category simply called nonmanufacturing. For example, in 1987 a sample of about 14,000 firms was selected to receive Form RD-I.

/7'

$180 $180 $140 $120 $100 $80 $60 $40 $20 $0

1986

1986

1990

1992

1994

1996

1998

2000

I-+- Real total _ _ Real nonmanufacturing I 2.I Totalandnonrnanufacturing R&D expenditures, 1986-2001 ($1996 billions).

,rom 1988 through 1991, about 1,700 of these firms were resurveyed, and ut 300 of the 1,700 were nonmanufacruring firms. {In the early 1970s, recognition by NSF that more detailed information on 'onmanufacturing R&D was needed began to spread. It was not until 1987, bwever, that NSF's annual R&D reports included R&D estimates separated trito three broad nonmanufacturing groupings: (1) communication, utility, ngineering, architectural, research, development, testing, computer .rogramming, and data-processing service industries; (2) hospitals; and (3) medical laboratories. By 1992, decisions were made at NSF to draw new "ampies annually with broader coverage, to increase the sample size from 000 firms to 23 000 firms, and to add 25 new nonrnanufacruring .industries to the sampling panel. Included in these 25 new industries were -.finance, computer and other business services, and engineering and manage.menr services. In 1992, the number of manufacturing firms nearly equaled the number of nonmanufacturing firms-ll,818 compared to 11,558. \'However, in 1993 and 1994, the nonmanufacruring firms sampled fell below the number of manufacturing firms sampled. . In 1995, the sample of nonrnanufacrurers was expanded by 60 percent. In that year, R&D expenditures for the following industries were, for the first time, reported in NSF's annual publication, Research and Development in Industry; transportation and utilities, including communications; trade; finance, insurance, and real estate; services, including computer-based

'd4

J.U~"''''H~Y'':

UC1UIIL1Ull:S VI

US NSF

J:\.Ol:U

EU Frascati

Thefollowing are indrtded in the definition of R&D: Basic research: Pursue new Basic reseercb is work

knowledge whether or not the search has reference co a specific application Limited to federal, university, and nonprofit organizations Applied reseercb: Apply existing knowledge to problems involved in creating a new product or process

Development: Apply

I

existing know ledge [Q problems involved in improving an existing product or process

I: I'

~

Ii

Table 2.2 Continued

US R&E tax credit bill

done to acquire new knowledge, withouc any particular application or use in view

Applied research is original

Research that is undertaken to discover information technical in nature and holds applications useful in developing a new or improved business component of the taxpayer Research that seeks a Dew or improved function performance, reliability, Or quality

investigation to acquire new knowledge directed toward a practical objective or a single product, operation, method, or system Experimental developNot specified ment is systematic work, drawing on existing knowledge aimed at producing new, or to improving substantially, existing products

R&D from acquired firms prior co acquisition Amortization above actual COSt of property and equipment related to firm R&D Test and evaluation once a prototype becomes a production model Routine product testing

Consumer, market, and opinion R&D; advertising new products or processes

Not specified

Not specified Not specified

Not specified

Routine product testing

General purpose data collection

US R&E tax credit bill

Geological and geophysical exploration activities Quality and quantity control

Analysis of soils and atmosphere Not specified

Not specified

Troubleshooting for breakdowns in production Social sciences, erc.: any research in the social sciences, arts, or humanities Nor specified

Scientific and technical information assistance Social sciences, etc.: any research in the social sciences, arts, or humanities Feasibility studies (e.g. a study of the viability of a petrochemical complex in a certain region) Administration and other supporting activities

Routine or ordinary testing or inspection for quality control Scientific and technical information assistance Social sciences, erc.: any research in the social sciences, arts, or humanities Efficiency survey

Activity relating to management function or technique

Source, NIST (2005).

Adaptation of existing business components to fit a particular customer's requirements Not specified

Not specified

EU Frascati

Management and organization R&D

The!o/!owi"g arenotinc/fJded in the deji'Jition of R&D: Not specified

US NSF

Research after commercial production of the business component Market research, testing, or development (including advertising or promotions) Routine data collection

/t

business services, health services, and engineering and management services; and other. In 1998, 19,973 nonrnanufacturing firms were surveyed compared to only 4,836 manufacturing firms.

Although NSF responded to underrepresentarion of the nonmanufacturing Sector in national R&D statistics, it has done so using Form RD-l, a survey instrument that was originally developed on the basis of an understanding of the innovation process in manufacturing (Link, 1996). By so doing, NSF was implicitly assuming that the innovation process underlying the expenditure of R&D is generally similar in manufacturing as in nonmanufacturing. And such an assumption underlies any attendant policies based on NSF's R&D data. That mayor may not be the case, Three different definitions of R&D, from three alternative sources, are presented in Table 2.2 to illustrate similarities and differences in that activity.

NSF identifies basic research, applied research, and development as the three distinct types of activities classified for reporring purposes under the rubric of R&D.3 All three activities require eirher the creation of new knowledge or a novel application of existing knowledge. Once a production process is established, activities associated wirh any further development of that process are not considered as R&D. Furthermore, NSF omits all social science research from its definition of R&D activities, as also shown in Table 2.2.

~'I.

I I

The Frascati Manual, which guides data collection throughout the European Union (EU) classifies R&D activities into three similar categories: basic research, applied research, and experimental development, and all activities classified as R&D must be in the pursuit of new knowledge or the discovery of new applications for existing knowledge, product, or process.

The Frascari definition lists relatively fewer disqualifying criteria for R&D activities. See Table 2.2.

Finally, the US 1981 R&E tax credit definition uses the term "qualified research" to identify activities aimed at creating new information or products and new applications of existing knowledge as applied to existing products.

The Act also addresses the issue of modification or adaptation, omitted by the other rwo institutional definitions. The Act states that modification or adaptation of existing products to meet a client's needs is not considered R&D. See Table 2.2.

Models of innovation in the manufacturing and service sectors Based on insights from four case studies of service-senor industries, summarized in Chapters 3 through 6, we illustrate a theoretical model of innovation in the service sector. first, for comparative purposes, we describe schematically innovation in manufacturing firms and then, building off that schematic model, we set forth our model of innovation relevant to service-sector firms.

Innovation in manufacturing sector firms One well-established model of innovation activity relevant to rechnoh based firms in the manufacturing sector comes from Tassey (1997, Figure 2.2 builds on the Tassey model; the figure illustrates di technology elements within the overall model of innovation. Each techi! element has a slightly different degree of public good attributes. ;', distinctions in terms of public good attributes make the model espec; relevant for policy analysis, but also useful as a benchmark for manufact . and thus a point of departure for modeling service-sector innovation.

At the root of the model is the science base, referring to the accumula , of scientific and technological knowledge. The science base resides in:,' public domain. Investments in the science base come from basic reseai primarily funded by the government and primarily performed globall universities, federal laboratories and some large companies.

Consider a representative manufacturing firm. Technology developm ': 10 the form of basic or applied research, generally begins within the fir'

2.q, â&#x20AC;˘

R&D laboratory. Technology development Involves the _application of scientific knowledge from the science base toward the proof of concept ofa new technology. Such fundamental research, if successful, .Ylelds a generic technology. If the generic rechnolngy has potential commercial value, followon applied research takes place toward development, and If successful, a

proprietary technology results.

Basic, applied, and developmental research occur. within ~ firm as a result of the firm's overall strategic planning. Strategic planning defines. the environment for entrepreneurial activities. And entrepreneunal acnvrry ... . influences production process development.

Entrepreneurial activity related to innovation ImplI~s different tl~mgs to different individuals, depending on their academJ~ or pr.ofessiOnal backgrounds. Therefore, we have reviewed in an append~x to ~hlS chapter the intellectual history of the concept of entrepreneurship as It relates to innovation. .' .

1 ¡ ure 2 .2 , we think of entrepreneurial acnvity 10 t re .

1n terms 0 f FIg Schumpeterian sense. Technology development corresponds to Schumpeters concept of innovation being, among other th.iogs, the ere-arion of a new g~d 'or. new ualiry of good or the creation of a new method. of productIOn. q .. .

I .. I rymg out of new :-;.1n this sense. entrepreneurial acuvrry IOVD ves rne car "'mbinations" (Schum peter, 1934, p. 74).'

"" . ¡n~.i~,~

-+

Production process development

----JI>-

r

Entrepreneurial activity

t

/"Technology development Proprietary

technologies

t

Generic

Market development

-+

New value-added product

<.

Risk reduction

t Infrastructure technologies

i 1 Sciencebase

technologies

[el-ofinnovation relevant to the manufacturing sector.

Infrastrucrure technologies, Or infratechnologies, support the processes that lead to both generic and proprietary technologies, and heoce technology development. Infrastrucrure rechnologies are a diverse set of technical tools thar are necessary to conducr all phases of R&D efficiently, Following Link ,I al. (l983b) and Tassey (1997,2005), examples of infrastructure technologies include measurement and test methods, process and quality control techniques, evaluated scientific and engineering data, and the technical basis for product interfaces. Infrastructure technologies thus influence the science bases and are so influenced by it.

The managerial skills necessary for a firm to move its proprietary technologies to a value-added product or process are also shown along the [01' horizontal area of the schematic in Figure 2.2. Afrer production, market development takes place. Markers do not always accept new technology for a number of reasons, including transaction COStS associated with verifying the new technology's attributes and interoperabiliry of the new technology with existing technologies. Infrastructure technologies can reduce such market risks and thus speedup market development. And to conclude, if market developmem is successful, value added will result.

Of course, all of the private-sector decision nodes in Figure 2.2 are influenced by the overall strategic planning of the manufacruring firm.

As shown in Table 2.3, the ten elements in the model in Figure 2.2 can be categorized in terms of their public good characteristics. This categorization is irnportanr because it highlights those aspects of manufacturing inn?vation that can be leveraged by public-sector innovation or technology poky. The knowledge that resides in the science base is a pure public good, as in knowledge per se. Generic technologies and infrastructure technologies are quasi public goods, and rhe results of risk reduction that

stem from the use of infrastructure technologies also have a public good nature since the innovating manufacturing firm cannot appropriate these results fully.

Innovation in seruice-sector firms Many of the fundamental characteristics of the innovation process differ between the model tel evant to manufacturing in Figure 2.2 and the model relevant to a service-sector firm in Figure 2.3.

Both manufacturing and service-sector innovation processes are driven by strategic planning. However, service-sector activities are more influenced by competitive planning because the technology-based service firm is more likely to innovate on the basis of customer input and on the basis of competitors that continually seek to challenge the firm for its customers. Whereas manufacturing firms srrategically formulate road maps for developing new emerging technologies, in the service sector, firms strategically formulate road maps for deploying modifications of existing products. Because manufacturing firms ate more likely to target discrete technology jumps, creating new technologies that make their competition obsolete, their strategic plans are long term and are linked less to current competitive planning. In contrast, service-sector firms' strategies are typically focused on retaining or gaining market shares and involve more continual or incremental rransirion strategies that are or will be integrated with competitive planning.

Both strategic and competitive planning drive the firm's entrepreneurial activity. Whereas entrepreneurial activity in a manufacturing firm was

StrategiC _ . pIanmng

Table2.3 Public good characteristics of the elements of innovation in the manufacturing sector

Elements of the innovation model Science base Generic technologies

Proprietary technologies Entrepreneurial activity

Infrastructure technologies Risk reduction Strategic planning Production process development Market development New value-added product

Pure public good Quasi public good

No No Quasi public good Quasi public good

No No No No

Product or service Market d --.. development enhancement an systems integration

i ~ .>

~~~~~~tive

Entr:~f~unaJ

Riskruction

Purchased

Purchased technical

Infrastructure

teChnOtl09;es

serv!iees

_

__

thnolor Science base

Figure 2.3 Model of innovation relevant ro the service secror.

Source: NIST (2005).

Enhanced value-added service

related to innovation creating a state of disequilibrium, in a service-sector firm the entrepreneur is more of a manager or perhaps an arbitrageur. In the sense of Baudeau (1910), the service-sector entrepreneur collects and processes knowledge and informarion. The identification and use of others' technologies are at the heart of entrepreneurial activity. R&D activity (subsumed under entrepreneurial activity in Figure 2.3) fulfills an adaptive role to ensure that purchased technologies and technical services are used efficiently. One could generalize about the service-sector entrepreneur as being an individual who allocates resources with an equilibrium state and to maintain an equilibrium state.

Whereas entrepreneurial activity, and its motivating perception of opportunity, drive the manufacturing firm toward producing new products and processes. the entrepreneurial activity of the service-secror firm drives redesigned or reconfigured enhancements of its existing products. At the root of entrepreneurial activity are others' intellectual capital and technologies that are licensed orpurchased to meet rhe firm's road maps for deploying modificationsof its exisring-producrs. Still, perception of opporrunity is rhe defin'.;"1'#g~;~ritt'efr~n·¥~iaFc~~cterjs·tib This product and service enhancement often ,.,,~n~~l¥'~;ts'yst~I#S"'~inl:e~i:ation· where systems integration facilitates the inter:.: ~;:~tib1i'8'frlfii:fdWa}~:J;sbftware; and the synthesis ofapplication domains such as "t¥iIin1'"i!f'?Hlit'fiili1i2Hirihg, transportation. and retail.

-'h.··,!t·:· ~"'@~,;T~'·:·1('fytdfsdnttion between manufacturing and service-sector firms' R&D is "rhar manufacturing firms conduce a larger share of their R&D in-house, and the output of that internal activity is more likely to be a proprietary technology. In the service sector, little research (R) occurs in house, and the development (D) activity rhar occurs is primarily related to enhancing, redesigning, or reconfiguring others' proprietary technologies.

Whereas manufacturing firms license or purchase others' technologies in rhe form of intellectual capital or equipment to be used to ptoduce proprietaty technology, service-sector firms purchase others' technology in the form of equipment to be modified and integrated into their operational system to deliver modifications to existing products. In addition) manufacturing firms strategically) through their research, introduce new technologically advanced products and processes to anticipate new consumer wants; whereas serviceseceor firms strategically, through information gathering, modify existing products to meet existing consumer needs. These differences underscore the differences between entrepreneurial activities in the two sectors.

Both manufacturing and service innovation are built on the scientific base of knowledge. The manufacturing sector is more likely to build on the science base directly or in collaboration with universities. In contrast, the service seceor purchases products and services as inputs that incorporate o..l-,ers' research, which of course draws on the science I, ~e.

~'f!.·;.,The role of infrastructure technologies is also different between the twO ')sectors. Whereas infrastructure technologies reduce the market risk sociated with the market introduction of a new product or process to the anufacruring firm, infrastructure technologies ensure that purchased techlogies interface or integrate with the service-seceor firm's existing terns. Such infrastructure technologies emanate from the science base, md it is the science base that is at the root of the production of purchased

ethnologies. ~-An important component of the innovative process in both the manufac.ng and service sectors is risk reduction. However, the focus of the ., ivities differs. In the manufacturing sector, innovation is likely to be less 'ntegrated with markering. Once a new product has been designed and ,.ested, technical risk may be relatively low, but market risk may be signifit because the product needs to be accepted and integrated into existing ystems. In contrast, service-sector innovation is more likely to involve enhancements to products in existing markets) lowering market risk. ;However, limitations and the cost of testing increase technical risk, making i~isk reduction a key objective of the product enhancement phase of service innovation. ',.'.' As part of their risk reduction strategy, it is not uncommon for serviceector firms to outsource key components of product developmenr or systems integration. However, many service-sector firms provide research as their primary service; thus, a component of their service is to assume the risk other .fi;ins are looking to outsource. These firms provide a key input into ~'ritrepreDeurial activities similar to purchasing technology embedded in 'products or licensing rechnology.? ':" In. summary, Figure 2.3 illustrates the process by which service-sector B.rms access and inregrate technology with rhe goal of developing and '~.iproviding enhanced services to their customers. These firms lead the strategic ".p lanning and entrepreneurial activities, as well as market development.

'They are likely ro be heavily involved in the final stages of developmenral -}iesearch but may oursource a large share of the applied research and early"'~tage developmental research. Their role is often an integrator of existing 'technologies; however, they may also oursource significant systems .. integration activities.j

Appendix The entrepreneur as innovator Throughout intellectual hisrory as we know it, the entrepreneur has wo~ many faces and "ryed many roles." Neither economic theory nor ecr -mic

Conclusions This case study represents an example of how one technology-intensive service industry can develop new services by acquiring new or emerging technologies and adapting them to fit into an existing service infrastructure.

In the case of a small network provider, technology is adapted and 'hen managed; feedback loops are established with OEMs to improve 'he networking products being manufactured. In the case of the larger network operators, innovation rests in the service firm articulating the development and deployment of the new service, bundling it with existing services.

In relation to the service-sector model of innovation presented in Chapter 2 (Figure 2.3), the service provider-in this case, the network operacor-sinnovation is taking place as the supporting infrasrructural technologies acquired from OEMs are brought together to create a system thar provides a novel or enhanced service. The network operators bear the majority of the product and service enhancement risk. Market risk, which is secondary. is in large pare passed on to aggregators.

4

Financial services industry

Introduction Over the pasr decade, changes in the US economy and the increased use of IT in financial services have significantly influenced innovative activities within the financial services sector. Increased competition through deregulation, consolidation, and disintermediation is forcing financial services firms ro innovate to maintain profitability.

At the same time, the demand for financial services has increased as the baby boom generation approaches retirement. The simultaneous increase in competition and demand is driving the industry to increase R&D to develop both new services and low-cost delivery devices for new and existing services.

The introduction of e-money, smart cards. e-checks, e-funds transfer, and improved encryption are some of the innovations developed in financial services in the twenty-first century. An additional trend influencing innovation is the disintermediation of financial services as manufacturers begin to encapsulate their products with services and deal directly with customers.

This trend is forcing service-sector firms to reduce costs by developing more efficient technologies and move into new product and service areas to mainrain profitabiliry.

This chapter profiles the financial services industry and its ongoing R&D expenditures and innovative activities. Findings from industry interviews are presented to provide an overview of the technology development and deployment process, with a focus on financial Web-based services.

Industry profile and R&D statistics Firms in the financial services industry, generally NAICS 52 and 53, consist of depository institutions; nondepository institutions; security and commodity

.J/

Case study of web services technology

brokers; insurance carriers; insurance agents, brokers, and services; real estate; and holding and other investment institutions.

Depository institutions include firms engaged in deposit banking and fiduciary activities. Nondepository institutions include firms engaged in extending credit in the form of loans but not engaged in deposit banking.

Security and commodity brokers, dealers, exchanges, and services include firms engaged in the underwriting, purchase, sale, or brokerage of securities and other financial contracts on their own account or for the account of others, and exchanges, exchange clearinghouses, and other services allied with the exchange of securities and commodities.

Insurance carriers include carriers of insurance of all types, including reinsurance. Insurance agents, brokers, and services include agents and brokers dealing in insurance and organizations offering services to insurance firms and policyholders. Real estate includes real esrate operators and owners and lessors of real property as well as buyers, sellers, developers, agents, and brokers.

Holding and other investment offices include investment trusts, investment firms, holding companies, and miscellaneous investment offices.

In 2003, total revenue for this secror exceeded $1.9 trillion wirh employment over 4.3 million. See Table 4.1 for aggregare and subindustry information. The financial services industry is tOO large and too diverse to examine in its entirety usi\ng interview-based tools. And much of the R&D performed by firms classified under NAlCS 52 and 53 is nor directly related to the provision of financial services. Please refer to this chapter's appendix for a discussion of the diversity of firms classified in the financial services sector. The remainder of this chapter focuses on financial firms classified under NAICS 523 as securities, commodities, contracts, and other financial investments and related activities.

Web services technology was identified for this projecr as a case study to demonstrate how innovation occurs within financial services firms. The firms interviewed are listed in Table 4.2. However, in a number of our interviews there was a shift in the discussion from Web services to more general concepts about R&D, how those activities are measured, and how they relate to the innovation process in general. The case example of Web services is provided in rhe appendix to this chapter.

The following discussion focuses on the financial services industry, in general, and R&D and innovation in the industry, in particular, in an effort to complement the information about the industry and relate to our model of innovation in Chapter 2. I

Defining financial services The firms interviewed defined financial services as an industry that provides services in, bur not limired to, retail banking, debt and asser management, and private and institutional investment. Financial services firms provide service to a variety of clients, ranging from individuals to large institutions.

Institutional client services include debt management, capital financing, public and private offerings of debr (srock and bonds), and equity, as well as other securities. The provision of these services is based on a skills set of underlying knowledge and experience to ensure high returns on investment for rhe client (consumer) and the financial services firm (producer).

Financial services change over time in response to consumer preferences.

Firms are continuously working to develop higher quality) more efficient, and Table 4.2 Firms interviewed related to Web services Name

Table 4.1 Summary of financial services industry (2003) NAICS Name 5222 5222 5231 5241 5242 5311 5511 5253

672 Depository institutions Nondeposirory institutions 99 Security, commodity broker 98 Insurance carriers 200 Insurance agems, brokers, services 43 92 Real estate 875 Holding and other investment offices 2,079 Finance, insurance, and real estate

Source: COMPUSTAT (2003).

Computer Service Firms Accenrure

Number of Employment Sales (thoUJamir) ($ mil/ions) firms 2,228 359 322 1,007 165 79 168 4,329

715,938 217,974 188,342 694,385 30,021 14,533 66,600 1,927,792

~:

,.,

.Cz..

Position oj the interviewee

Nireo Partners

Associate parmer for financial services group Project manager

Financialservice institutions Merrill Lynch JP Morgan Chase Wachovia

Chief technology architect VP of treasury security services VP of retail integration

Industry research firm Forrester research

Principal analyst

Source: NIST (2005).

less expensive products that will allow them to gain competitive advantage in their inclusrry or sector. To this end, technology is developed andlor acquired and used to increase the productivity of the existing service and to develop new services. The developmenr and application of intellecrual capital (IC) is the dtiving force behind innovation.

Financial services can be either labor intensive, relying heavily on personal interactions and human capital, or capital intensive, relying on automation to lower the cost of transactions and the dissemination of information.

There are two general categories of these services: investment services and retail banking. The distinction between these two segmenrs lies in the level of technology needed to meet consumer demand for services in the respective segments.

Investment services rely heavily on a number of different technologies to ensure that accurate and com plere information is available to investors. A high rate of innovation is required as the leading firms compete for customers through differentiation via technology. We found that investment services firms were conducting a large share of innovation in-house. The largest investment services firms even have teams referred to as Advanced Developmenr Groups (ADGs) that are responsible for innovation and development projects within the firm.

Conversely, retail banks see themselves as competing for customers through the quality of service rather than the development of and innovation within services. Retail banks believe that they have found an optimal distribution between maintaining and enhancing the level of technology facilitating the provision of retail banking service. In recent history, the trend in the provision of services has been to move from human interaction toward total automation via the Internet and ATMs. In some cases, banks began to charge customers for using bank branches and human tellers to make transactions. However, recent literature suggests that this trend is reversing.

Although retail banks are maintaining the use of the Internet and ATMs for simple banking services, such as balance inquiries, deposits, and withdrawals, these banks are relying more on face-co-face interactions when providing more complex services) such as mortgages and mutual fund investments.

R&D activity The NSF's definition of R&D activities includes the following components: planned, systematic pursuit of new knowledge or understanding toward general application (basic research); acquisition of knowledge or understanding to meet a specific, recognized need (applied research); and, application of knowledge or understanding toward the production or improvement of a -oducc, service, process, or method (development). ;-'1'

,.".

II 'J'

~ ~'

.

".; '~.

Based on our interviews, we found that these concepts of R&D do not resonate well wirhin rhe financial services industry. None of the firms interviewed indicared that they conduct activities in eirher basic Or applied research. Most of the individuals interviewed could relate to the concept of development; however, rradirional research, either basic or applied, was not commonplace in their operations. We found that the ideas underlying the developmenr phase of R&D resonated best for investment services institutions. However, only the largest retail banking firms thought that they were performing development activities, and the smaller retail banks indicated they were not doing any R&D. For large investment firms, development activities mentioned in the interviews included conceptual design, articulation of technical specifications and capabilities, integration, and deployment of new technology. The research activities of large retail banks are primarily related to initial conceptual design, with only a small role in the developmenr and deploymenr of the technology. Table 4.3 identifies and compares development activities performed within investment and retail service firms.

Bullet marks in Table 4.3 indicate the stages in the innovation process where the financial services firm or the external technology vendor is conducting development acrivities. This table highlights the fundamental difference between investment services and retail banking in terms of development activities performed in house versus those performed by vendors. Large investment services firms conduct most of their development activities inhouse, relying on external technology vendors only for existing technologies that meet capability requirements and for assistance in implementing the purchased technology. Retail banking firms perform very few development activities in-house, relying on off-the-shelf rechnologies developed by vendors.

Table 4,3 Differences in investmem service and retail banking development strategies Ltmouation process

I messment services

In-home

Generation of new technology idea Initial development • Sources of generic technology Technology

Vendor

Retail bcmking Consortiums

In-bouse

•

•

implementation

Operation and maintenance Sowee: NIST ('-'1)05).

•

COlJJortiU1lJJ

•

infrastructure

Modification and

Velldr.tr

•

Our interview information suggested that different types of financial services require different levels of technological innovarion. In addition, a firm that is considered a market leader may differentiate itself from other competing firms through the continued development of new products or services. Smaller industry participants are reluctant to compete in innovation and instead find it more cost-effective to adopt off-the-shelf innovations from vendors.

Retail banks conduct very little of the development activities in-house.

However, based on our interviews, investment firms spend less than 20 percent of a project's total cost on purchased technology from external vendors. The majority of the costs are incurred through in-house activities relating to the adoption and modification stages of this innovation model.

The following are examples of innovation from our interviews in retail banking:

•

The innovation process The innovation process is a term used to describe the continuum of activities associared wirh how firms or industries develop and deploy new ideas and technologies. This process has long been discussed in the academic and policy literatures with respect to the manufacturing and industrial sectors of the US economy (e.g. research, prororype development, and scale-up to full production).

However, we found the same process does not characterize innovative activities in the financial services industry. Innovation in financial services seems to be driven by customer demand, where only the larger firms participate in innovative activity and that activity is a strategic response to compete for customers, as is the case in manufacturing firms. Smaller financial services firms do innovate but generally not in terms of enhancing state-of-the-art consumer services. Rather, smaller firms innovate by providing consistency in the level of customer service.i Larger financial services firms are conducting activities that conceptualize aod develop technological advancements. Retail banking firms innovate by adopting and somewhat modifying existing technologies from vendors. Large investment firms either develop technologies in-house or purchase technologies from vendors. Large firms then significantly modify vendors' technologies to meet their system needs. Smaller investment firms, on the other hand, could be considered imitators. They adopt technologies developed by external vendors, similar to the approach taken by smaller commercial banks, and use that modified technnlogy to provide a differentiated product.

The innovation process for financial services firms begins in-house with the following steps:

•

:f:

The following are examples of innovation from our interviews in investment services:

• •

• • •

The firm identifies a new service product to meet an actual or perceived. cusrorner need.

In-house development occurs CO specify the attributes of the needed innovative technology that will support the new or enhanced service.

The firm contracts an IT specialist to build the new technology.

The firm incorporates the purchased technology into its business process.

I~

ATMs and the Internet: Retail banks incorporate Internet capabilities into retail banking ATMs by purchasing ATMs from vendors and then incorporating Internet capabilities (in-house) to allow customers to perform online activities, such as bill paying and financial transfers to third parries. The distribution of development COStS for this service was 40 percent to pnrchased IC and 60 percent to in-house adoption and modification. Purchased IC is embedded in ATMs from vendors and the application of existing technologies relared co Internet Web services.

In-house activities include the large amounts of computer programming necessary to synchronize the ATM's Internet-based transactions so that the customer's accounts are updared automatically.

Web Services (see the appendix for detailed discussion): this area involves the adoption of Web service technology to increase efficiency in inrrainsritutional banking. This innovation is taking place through a collaborative cost-sharing project at the FSTC in cooperation with NEC and Stanford researchers. The distribution of development COStS was 80 percent to purchased IC and 20 percent for in-house activities. In the case of Web services, only a small amount of development was done in-house.

Purchased IC is embedded in technical consulting services from Niteo Partners. In-house activities include the bank's time spent overseeing the project and offering insight into the business practices the consultants were trying CO model in their Web services applications.

', I:

Virtual deJktop: This technology allows a financial firm's employees to log in to an institution's internal system via the Internet to conduct business or modify documents. The innovation is that the virtual desktop will not lose data if the user's connection is terrninared. This means that an employee working off-sire can log in to a firm's internal system, work off of shared documents that are housed on a server, terminate his or her connection, reconnect, and continue working on rhe Same document J~

without losing any information. The distribution of development costs for chis example was 5 percent co purchased IC and 95 percent for in-house development activities. Purchased IC is embedded in the compurer software platform developed by a vendor. In-house activities include performing tasks such as writing code, integrating the information format from the vendor with the in-house code, testing, and implementing

â&#x20AC;˘

the technology.

Virtllalj/rivate networks (VPNs): Aimed at addressing the security issues associated with connecting and conducting monetary transactions over the Internet, this technology ensures that the customer is able to access all or almost all of the firm's internal information system from a remote location as if he were accessing the system on site or from within the firm's nerwork firewall. The distriburion of development costs for this example was 20 percent co purchased IC and 80 percent co in-house development activities. Purchased IC is embedded in networking consultancy services specializing in network security. The consultant was responsible for developing the security code specification that ensured the security of information shared between the financial institution and the customer over the VPN. In-house activities include the articulation of specifications and capabilities that the VPN needed co meet. After the consultant built the VPN, the financial institution took the system and integrated it into the institution's existing line of service products.

R&D metrics for financial services The NSF has traditionally reported the ratio of R&D co sales as a metric co characterize the invesrment innovation intensity of manufacturing firms.

Academics and policy makers have similarly relied on this measure, and the measure is one that aptly characterizes rnanutacruriug's view. From a policy perspective, maximizing the R&D intensity of firms is viewed as a positive (and meaningful) objective co achieve growth.

However, our interviews indicate that this metric may be less useful for firms in the financial services sector. Firsr, R&D is not a generally accepted descriptor of innovative investments, and second, no individual in financial services spoke of maximizing innovative investments as being associated with growth. Growth, in the traditional paradigm, comes from in-house development of proprietary knowledge that is either cost reducing or product enhancing. In fact, whereas innovation in manufacturing is often cost reducing, that concept is orthogonal to strategic planning in services. ~emant.ic~.ly, financial services firms do not call this R&D, but compared with acnvmes

r"i

that occur in manufacturing, this activity is the same in nature as what many manufacturing firms call R&D. For the financial services sector, a more relevant metric proposed was the ratio of dollars spent to maintain existing technology versus the dollars spent to enhance or purchase new technology. An industry rule among larger retail banking institutions is 70 percent maintenance and 30 percent enhancement, or 2.3 to 1. If a firm moves closer to 80 percent maintenance and 20 percent enhancement, or 4 ro 1, that firm would not view itself as competing successfully for cusromers. Small retail banks do not follow this metric because of their strong reliance on prepackaged software and off-the-shelf solutions.

The 30 percent in enhancement or innovation is then optimally (i.e. in a cost-minimizing manner) allocated between in-house development activities and the purchase of IC from vendors. This rule could be considered the optimized innovation ratio for financial services firms.

In investment services, an efficiency measure in technological innovation was suggested as the best way to evaluate a firm's level of innovariveness.

This metric measures technological success through a ratio of the cost of technology to the firm's revenues. The cost of technology represents the sum of purchased intellectual properry from outside the firm and in-house development activities. Holding the quality of service constant, the firm's goal should be to minimize the ratio of its total technology-related expenditures to services. As part of meeting this goal, the firm minimizes technology costs by optimizing the ratio of in-house development activities to the purchase of IC from veodors.

To illustrate the in-house development versus IC purchase decision, consider Figure 4.1. It depicts what we call an iso-rechnolcgy curve, To. The vertical axis represents purchased IC. The horizontal axis represents in-house IC Purchased IC generally cakes the form of purchased equipment and labor including human capital, and in-house IC generally cakes the form of human capital. For the financial services sector firm at point A. corresponding to a given level of purchased and in-house IC, innovation by the firm could be described as an outward shift in To to T I , where T 1 represents a new bundle of services to meet customer needs.

A firm maximizes technology output by selecting the most cost-effective combinarion of in-house to purchased IC As rechnology demands for new products and services increase, the optimal pathway for service firms may diverge from manufacturing firms because of differences in core capabilities and business models. Although we did nor collect financial information from those interviewed, it was our impression that a greater percentage of IC in service firms came from purchased technology than from in-house

.r-)

PIC,

1 .: Technology

!

Purchase PIC j intellectual capital

pathway ,: (Engle curve) "::"

- - T,

(PIC) PICa

"'---~c:_----

IHIC,

IHIC,

T,

IHIC,

In-house intellectual capital (IHIC)

Figure 4.1 Iso-technology curve.

Source: NIST (2005).

technology. And this trend is likely to increase as systems become more complicated, resulting in the curvature of the technology pathway.

Examples of R&D activity in security and commodity broker services Minrel Internarional Group Ltd_3 suggests that the financial services industry is in the midst of redefining its business processes by developing and adopting technologies such as data warehousing and mining, customer service and support software, and client relationship management (CRM) tools, The industry's goal is to provide either the individual or the corporate investor with easy and reliable access to real-time market investment information.

These services are being provided via networks that require higher levels of security co ensure the confidentiality of monetary transactions.

In addition [Q improving the services for customers, financial institutions are also interested in developing technology for investment tools that will enhance a financial analyst's ability [Q make the most lucrative investment decision and gain competitive advantage. Industry leaders have focused on developing databases that update information in real time, thus equipping

s:

investors with up-to-the-minute market information and assisting investment managers in decision making. The following examples describe the rypes of R&D activi ties performed by financial services firms.

Financial markets can be extremely volatile in both the short run and the long run, and without a way of providing up-to-the-minute market information, investment managers' firms may realize large losses. To this purpose, industry leaders in investment services have turned to innovation as a strategy to improve process quality and acquire additional market share.

Morgan Stanley has developed real-time international hedge funds and equity indices informed by market data via Reuters and Bloomberg. Morgan Stanley will be able to offer emerging market, regional, country, and sector equity indices in real time. Their real-time indices are sa.id to provide a unique insight into the intraday movements of the global equity markets and thus enable clients to evaluate their portfolios' performance versus the benchmark Morgan Stanley index at any point in the day.

JP Morgan's Investor Services produce development division announced in October 2002 a strategic alliance with Investors, a leading supplier of performance measurement services for equity research. The collaboration goal is to develop an integrated research and benchmarking tool that will help institutions maximize returns through better analysis of their supplier networks and relative contribution to perforrnanre OP Morgan, 2002).

Benchmarking tools are a way for financial institutions to demonstrate the value added from their research and evaluate comparatively the performance of research analysts in-house against other institutions.

Merrill Lynch decided to enhance their services by developing a superior plarform for their financial advisors. In November 2002, they partnered with Thomson Financial, a unit of the Thomson Corporation, to develop a Wealth Management Workscation (WMW). The workstation will be designed [Q support financial advisors through the use of robust market data, news and portfolio management tools, and CRM software.

In addition to product development efforts using partnerships as mentioned above, broad research consortiums are formed to conduct and coordinate generic and infrastructure R&D. Inreroperability issues can be addressed through the cooperative efforts of member organizations, with individual institutions realizing gains in competitive advantage through the development of proprietary technology, For example, the Financial Services Technology Consortium (FSTC)' is a member organization of the leading financial firms in North America. The FSTC coordinates collaborative technology research and development through pilots, proof of concept, tests, and demonstrations [Q develop interoperable, open-standard technologies that answer core competency needs for industry. FSTC prototypes new

6"r

infrastructures for financial transactions, confirms new specifications for the industry, and evaluates new technologies in lab settings.

The FSTC has conducted projects since 1994 in areas such as customer authentication, branch automation, check truncation, Web services, wireless banking, and biometrics. By developing these services through cooperative effaces of the organization's members, the technology developed ensures open architectures and inreroperabiliry, The financial services industry identified the need for technological innovation in the areas of voice authentication, Web services for corporate cash management, automated intrainstitutional exchange systems, and electronic checking. The following are some examples of the type of projects performed by the consortium's working groups.

Universal value exchange (UVX) is a set of protocols that define the internal architecture, interfaces, and gateways to existing payment systems for financial institutions. Thought to be middlewere, the UVX protocols will support payment processing such as papet check processing, wire transfer services, ACH, ATMIEFT, and credit card processing. Initially, the architecture is planned for transactions between banks and patrons; however, following the adoption by two or more banks, UVX is designed to conduct inrrainstirutional transactions. The technology goal is to reduce the operational costs of legacy payment systems by connecting existing systems to a modern payment infrastructure using XMl, state-of-the-art security technology, and current Internet protocols.

The ANS X9.85 project will test the viability and performaoce of a prototype check validation program by designing a system that uses the ANS X9.85 standard "Specifications for Automated Identification of Security Features." The system will be evaluated using a metric developed by financial institutions. The goal of the project is to surmise the degree of difficulty in modifying current check-processing systems to identify security features, assess the scalability and ease of integration with existing systems, and specify technical and operational barriers to implementation.

Conclusions This chapter demonstrates the diverse and abundant examples of R&D and innovation activities conducted in the financial services sector. The financial services sector has used IT to enhance, its existing services and, in some cases, to even create new services. In addition, we have seen that these firms conduct R&D activities both collectively and independently in their efforts to innovate. Furthermore our interviews demonstrate that in highly competitive service industries, increasing market share is a major driver of

innovation. Service firms use innovation as a strategy co differentiate their services in a competitive marker.

In relation to the service sector innovation model presented in Chapter 2 (Figure 2.3), the financial service provider has engaged in a variety of cost-sharing strategies that use purchased IC, software, or other technologies to articulate, develop, and implement innovation in the form of new or enhanced services. Because information systems are complex and tend to be custom built, much of the systems development (and hence the risk) occurs in house, although financial services firms actively pursue risk reduction by participating in consortia where inreroperabiliry issues are addressed through prototype platforms and test beds.

Appendix

Diversity in the financial services sector This appendix provides additional background related to innovation in the financial services sector. Table 4A.l lists the 10 largest financial services firms, ranked by sales. Table 4A.2 lists the publicly traded firms for which sigoificant R&D expenditure information is available. As shown in Table 4A.2, most of the firms with significant R&D expenditures are either holding or technology investment firms and are not necessarily research-related firms conducting R&D for the provision of financial services. For example: â&#x20AC;˘

â&#x20AC;˘

SEI Investments Company is a consulting firm specializing in financial management and investment technology solutions. Although rhe firm is oriented around the financial sector, the R&D reported seems to be emerging from the development of software applications that enable clients to make decisions concerning their investment portfolios.

Anglo American PLC-ADR, with the second largest reported R&D expenditure, is a holding company, controlling the majoriry of shares for some of the world's largest diamond (45 percent of DeBeers), gold (53 percent Anglogold), and platinum (50 percent Anglo American Platinum) firms. Also Anglo American PLC-ADR is one of the world's largest independent coal miners, with interests in ferrous and base metals, industrial minerals, and forest products. Only $4 million, or 0.17 percent, of the total $2.4 billion operating profit for Anglo American was dedicated to financial services in 2001. yet they reported a $34 million investment in R&D. Given the brief firm description, it is unlikely that the amount invested in R&D is being used to innovate in the financial sector.

6'1

Table 4A.l Ten largest firms in the financial services industry (2003) Name

NAiCS cod,

Employment (thousands)

Sales ($ mil/ions)

Cirigroup Inc. Prudential PLC-ADR Bank of America Corporation ING Groep NV-ADR American International Group Fannie Mae

5223 5241 5221 5241 5241 5222 5231 5231 5221 5241

115.0 23.0 155.9 82.7 55.0

82,005 51,745 51,526 43,819 40,656 36,968 34,879 33,928 33,544 26,959

Merrill Lynch & Company Morgan Stanley Dean Witter Chase Manhattan Corporation Allstate Corporation

NA

67.2 55.3 74.8 52.0

Source, COMPUSTAT (200,)).

Note NA: nor available in COMPUSTAT. ~o

NN

IGEN Inc. is another example of a firm whose producr is nor related ro rhe financial sector. IGEN designs and manufactures diagnostic systems that aid in the mapping of the human genome. It holds the patent rights foe this cutting-edge technology and leases its use to clients such as the Human Genome Project and other molecular biologists. The firm's Standard Industrial Classification (SIC) code, 6794, classifies the firm as a patent owner and lessor.

The firm's description demonstrates that the R&D reported ro COMPUSTAT is advancing medically related service industries, not the financial sector.

MIPS Technologies is a design firm, classified as a patent owner and lessor, specializing in developing the low-power 34 and 64 bit core chips that are found in most video game consoles in roday's markets. In addition to gaming system chips, they offer microprocessors and architecture design systems.

Instead of manufacturing these products, they Simply license their intellectual property ro large manufacturers ijfhigh-tech products, such as Hewlett-Packard, NEC, and Philips Semiconductors.

Rambus provides chip and system firms with interface solutions to enable high performance and system bandwidth for a range of consumer, computing, and networking applications. Rambus provides its customers with interface solutions and comprehensive engineering services to support implementation of irs interfaces in customer products.

Example of R&D activities in selectfinancial services sector firms These examples provide evidence of the breadth of acnvmes that are ""oounted for in the financial sector. Table 4A.3 list' ~rms reporting R&D

~y

~o

r- r-

~

on

cici

N

on

cici

Table 4A.3 Select sample of firms reponing R&D expendirure in the financial services industry

Name

NATes

Investment Technology Group Inc.

5231 5231 5231 5223 5223

A B Wadey Group Inc. IUFE.com Inc. Mortgage.corn Inc. LendingTree.com Inc.

Employment Sales R&D (t"o"sands) ($ millions} expenses ($ millions) 0.3

0.1

NA 0.5 0.1

232 21 12 43 7

9.7

6.0 3.0

2.9 1.1

Source: COMPUSTAT (2003).

Note NA;noc available in COMPUSTAT.

expenditures related to the financialsector. For this nonrepresentative sample of firms, information on their Web sites revealed the following innovation-related informacion.

Investment Technology Group Inc. is a full-service rrade execution firm that uses technology to increase the effectiveness and lower the COSt of trading by emphasizing R&D in the products they offer, which are as follows: POSIT: an electronic stock crossing system; QuantEX: a Unix-based decisionsupport, trade management, and order routing system; SmartServers: serverbased implementation of trading strategies; Electronic Trading Desk: an agency-only trading desk offering clients the ability [Q efficiently access multiple sources of liquidity; ITG Platform: a PC-based order routing and rrade management system; ITG ACE and TCA: a set of pre- and posrtrade tools for systematically estimating and measuring transaction costs; ITGI Opt: a computer-based equity portfolio selection sysrem; ITG WebAccess: a browser-based order routing tool; and Research: research, development, sales. and consulting services."

A.B. Watley Group Inc. is a New York-registered broker-dealer thar operates both direct-access trading and third market institutional sales rrading brokerage businesses. The firm offers a proprietary technology called Direct-Access Vertical Exchange (DAVE)" ro brokerage and banking industries. DAVE consists of a ticker plant, order entry and trade processing, and data delivery engines.

LendingTree.com Inc. is a lending exchange that attracts customers looking for loans and processes loan requests through a number of banks.

LendingTree licenses their technology platform LEND-X(SM) (which powers their Inrernet-based lending exchange) to other businesses to create exchanges on their own Web sites.

Web services case study Consortiasuch as the FSTC serve an important role in technology development activities, and they contribute significantly to applied research being performed to support the financial services industry. The following is a case study of the proof-of-concept project performed by a group of retail banks in cooperation with NEC and Stanford University through rhe FSTC.' We found little evidence that retail banks were involved in applied research other rhan through funding of consortium activities. Our interviews indicate that NEC performed all basic and applied research and thar only a small percentage of the total project cosr was spent by banks for in-house development activities.

The FSTC recently completed the proof-of-concept project relared ro Web services. The project's goals were to promote shared learning and develop technologies related to Web services for identification, aggregation, and composition of corporate account data and services. The Web services project was cosponsored by NEC's Nireo Partners and three retail banking firmsWachovia, Bank of America, and JP Morgan Chase.

Niteo Partners is a wholly owned subsidiary ofNEC (NEC has worked on innovation in technology for over 100 years). NEC has a large investment in technology research labs around the world. These laboratories work on issues related to Internet software, nanocornputing, quantum cryptography; and other nerworking-related technologies. In addition to applied rechnology research) NEC also invests in academic or basic research. NEC has worked closely with a multidisciplinary research faculty at Stanford University.

Areas of research include knowledge representation, machine-to-machine communication and interaction, and automated computing.

The NEC invesred $2 to $3 million for academic research in machine-tomachine interactions and wanted to set up a project that would advance awareness of its technology. NEC was looking for market exposure in the financial services secror specifically. NEe called on its wholly owned subsidiary) Niteo Partners, to design or craft a project in conjunction with the Stanford research team that would demonstrate the machine-co-machine interaction technology in combination with existing standards related to Web services to reduce the COSt of intrabank interactions.

The goals of rhis project were '0 gain market exposure for NEC technology related to Web services and to demonstrate what Web services have the potential to provide in the future for the financial sector. .

When Niteo approached the FSTC, they discovered that the financial services sector was just beginning to understand the current capabilities and applications for Web services. Niteo spent 6 months educating industry

participants on the state of standards related to Web service applications.

Nireo then created case examples to demonstrate the market potential for Web services in the financial services sector.

Once the returns on investment were visible, retail banking institutions entered inro a cost-share project with Niteo through FSTC. This project allowed industry experts ro share knowledge of the business and specify the special needs that the technology needed ro address. Niteo used this input to develop reference materials that could be implemented in-house by the participating banks.

The latter project was cofunded by NEC and the FSTC participants, including Wachovia, Bank of America, and Jp Morgan Chase. These firms agreed ro fund the proof of concept jointly in return for the rights ro core findings and any intellectual property developed. Nireo ran the project. The knowledge and learning occurred at Nireo but was informed by rhe business expertise of the banking executives from the participating banks. The project has moved into a second phase concerned with implementing rhe findings from Phase J. Phase 1 lasted approximately 2 years.

As parr of the implementation phase of rhe projecr, parricipating banks have the opporruniry to become more involved in developing and deploying the technology. At this poinr Jp Morgan Chase is rhe only one of rhe rhree banks that is proceeding wirh implementation. They are funding rhe implementation of the basic standardized codes developed as parr of Phase 1 and are creating a test bed to work through the remaining technological issues.

Most of rhese acriviries will be conducted in-house, with NEC working as a consultant. Insights gained in Phase 2 will be IC owned by Jp Morgan Chase and will not be shared with the other banks that participated in the first phase of the project.

5

Systems integration services industry

Introduction Systems integration services facilitate the intersection of hardware, software, and pragmaric indusrry knowledge that provides rhe foundarion of IT systems.

The Computer Science Telecommunicarions Board (CSTB, 2000) defines sysrems inregrarion as rhe wiring together, via hardware and frequently very complex software, of the often already existing islands of computer applications into a coordinated enterprise-wide distributed network system. Systems inregrarion includes more than just physically allowing incomparible componenrs to communicate. It is the synthesis of application domains such as finance, manufacturing, transportation, and retail and the supporting information infrastructure including databases, operating systems, architectures, networks, communications devices, and security measures (CSTB, 2000).

Systems integrarion industry Since 1992, systems integration has been the most rapidly growing component of the US computer industry. Total revenues for custom integrated system design and custom programming services rose from $34 billion in 1990 to $76 billion in 1997 (CSTS, 2000).

Table 5.1 illusrrates rhe diversity of companies providing systems integration services and highlighrs rhe variance in their R&D reporting.

Service-only firms are generally less likely to report their systems integration activities as R&D expenditures. In contrast, diversified firms such as IBM and Hewlett Packard classify one-quarter to one-half of systems integration acrivities as R&D; thus, the level of reported R&D spending by diversified service firms is significantly greater than that of service-only firms.

This variance in reported R&D for systems integration is likely due to several factors. Larger systems integration firms differentiate themselves

,

"