MASSACHUSETTS INSTITUTE OF TECHNOLOGY SLOAN SCHOOL OF MANAGEMENT 15.565 Integrating Information Systems: Technology, Strategy, and Organizational Factors 15.578 Global Information Systems: Communications & Connectivity Among Information Systems 1 Spring 2002 Lecture 22 DEVELOPMENT OF STANDARDS: FOCUS ON DATA STANDARDSISSUES • WHY NOT JUST ONE BIG ORGANIZATION? -- BENEFITS OF LOOSELY-COUPLED ORGANIZATION -- BOUND RATIONALITY <----> UNCERTAINTY / COMPLEXITY • THUS, NOT ONE BIG PIECE, LOTS OF PIECES. -- HOW DO WE COORDINATE? • STANDARDS PLAY A KEY ROLE 2STANDARDS • WHAT TYPE OF ISSUE? -- TECHNICAL -- STRATEGIC -- ECONOMIC 1 -- ORGANIZATIONAL2 (POLITICAL) • HOW ARE INTER-ORGANIZATIONAL STANDARDS DEVELOPED? -- INDUSTRY-WIDE • WHAT ARE INSIGHTS FOR INTRA-ORGANIZATIONAL STANDARDS? -- INTERNAL TO A COMPANY 1 Marshall Van Alstyne, Erik Brynjolfsson, and Stuart Madnick, "Why Not One Big Database? Principles for Data Ownership", Decision Support Systems, Vol 15, December 1995, pp. 267-284. [SWP #3695, CISL #94-03] 2 Marvin Sirbu and Steven Stewart, “Market Structure and the Emergence of Standards,” CMU Working Paper, April 1985. 3DATA COMMUNICATIONS STANDARDS • FIRST DATA COMMUNICATIONS STANDARD -- PARIS, 1865, INTERCONNECTION OF TELEGRAPH SYSTEMS • X.25 PACKET SWITCHED NETWORK -- SIRBU CASE STUDY • PACKET SWITCHED NETWORK DEVELOPMENTS -- USA TELENET (1974); FRANCE TRANSPAC (1973); CANADA DATAPAC (1974), ... • CONTROVERSIES / COMPROMISES (recall ATM 53 bytes standard) --X.25 CONTROVERSY OVER DATAGRAM VS VIRCUIT CIRCUIT -- RESOLUTION: PROVIDE BOTH (DATAGRAM RARELY IMPLEMENTED) -- PACKET SIZE DIFFERENCES: GERMANY (128, 1024), FRANCE (64, 128), JAPAN (32, 64, 128, 256), ETC. -- PACKET SIZE (128 BYTES COMMON), RETRIES, ETC. 4 CCITT USA UK - AT & T + TELENETX CANADA -+ DATAPACX FRANCE + TRANSPAC CTCAMODEL FOR STANDARD DEVELOPMENT PREDICTION • TYPES OF COMPATIBILITY -- PEER-TO-PEER: TWO FACSIMILE MACHINES -- INTERFACE: TV AND VCR • VENDOR -- CENTRALIZED (INTEGRATED) -- SUPPLIES ALL COMPONENTS -- DECENTRALIZED (FRAGMENTED) -- NO ONE SUPPLIER PROVIDES ALL -- DIFFER IN EXPERTISE, -- PATENT/ LICENSING -- GEOGRAPHY -- CAPITALIZATION • BUYER -- CENTRALIZED -- ONE PARTY BUYS ALL COMPONENTS -- DECENTRALIZED -- UNRELATED PARTIES -- MODEMS, AM TRANSMIT/RECEIVER 5SITUATIONS • MAJOR FACTOR IS COORDINATION COSTS 1. CENTRALIZED PURCHASE, CENTRALIZED PRODUCTION -- MULTIPLE STANDARDS CAN CO-EXIST (NO STANDARDS) -- NO TRANSACTION COSTS -- CANNOT SWITCH SUPPLIERS 2. DECENTRALIZED PURCHASE, CENTRALIZED PRODUCTION -- STANDARD IMPORTANT (FACSIMILE TO ARBITRARY PEOPLE) 3. CENTRALIZED PURCHASE, DECENTRALIZED PRODUCTION -- NO ONE FIRM SUPPLIES ALL COMPONENTS -- STANDARDS, BUT NOT UNIQUE (MULTIPLE STANDARDS) (CASSETTE AND CD; MULTIPLE OPERATING SYSTEMS) 4. DECENTRALIZED PURCHASE, DECENTRALIZED PRODUCTION -- STANDARDS MOST IMPORTANT (X.25, TV) 6SUMMARY OF KEY SITUATIONS PRODUCTION PURCHASE DECISION (BUYER) DECISION (VENDOR) CENTRALIZED DECENTRALIZED 1. NO 2. STANDARDS IMPORTANT CENTRALIZED STANDARDS - CPU/DISK - DIAL-UP MODEMS - LOCAL E-MAIL - FACSIMILE 3. MULTIPLE 4. STANDARDS VERY STANDARDS IMPORTANTDECENTRALIZED -PBX -TV - LAN - PACKET NETWORK 7DATA STANDARDS / INTEGRATION: EXAMPLE AND COUNTER-EXAMPLE 1. PART_NUMBERfor 3/4” SCREW 2. CUSTOMER_IDCodes for IBM 3. Definition ofSALES Integrated Non-Integrated Div A Div B Div A Div B 115899 115899 115899 718 42765 42765 Adjust Adjust returns returns 42765 42675, 49345, etc. Adjust Not adjust returns returns Example: Venezuela Nationalized Oil Companies – drilling pumps 8Data Standards and Integration Research The Impact of Data Integration on the Costs and Benefits of Information Systems Data integration 1. Ability to share corporate-wide data to 2. Flexibility to respond to systems Costs and benefits of information systems ( + ) ( - ) ( +, - ) address subunit interdependencies subunit needs for locally unique information 3. Costs of designing and implementing information • Note: tradeoffs might be different for each sub-unit. • Most successful if sub-units: (1) interdependent and (2) not highly differentiated. Source: Dale L. Goodhue, Michael D. Wybo and Laurie J. Kirsch, “The Impact of Data Integration on the Costs and Benefits of Information Systems”, MIS Quarterly , September 1992, pp. 300. 9Lessons from the Field Organization Outcome Ventura SSD Textbook success story Ventura Finance Turning point in divisional support for IS LSA Still struggling to get organizational buy-in Cedar De-emphasized effort Derrick Shelved after reorganization Nat. Tech. Shelved Consumers Shelved Foothill Shelved Source: Dale L. Goodhue, Michael D. Wybo and Laurie J. Kirsch, “Strategic Data Planning: Lessons From the Field”, MIS Quarterly , March 1992, pp. 23. 10Benefits from Integrated, Sharable Data + • Improve communications across subunits (view for top mgt) – Data integration facilitates the collection, comparison, and aggregation of data from various parts of the organization, leading to better understanding and decision making when there are complex, interdependent problems. Devlin Electronics: Issues: They found a number of interlocking problems: lant capabilities, and plants Outcome: corrective action. ML On-time deliveries (a critical competitive issue in the semiconductor market) fell to only 70 percent, a multi-disciplinary team at Devlin Electronics used organization-wide integrated scheduling data to track how production schedules were made, changed, and adjusted by the many different groups involved. some plants were not properly updating their inventory levels and equipment conditions, marketing was overriding the schedule without regard to pwere overriding the system without regards to plant capabilities, and plants were overriding the system without regard to critical order requirements. Organizational-wide integrated data allowed Devlin to understand its problems and take On time delivery improved from 70 percent to 98 percent Merrill Lynch (ML) - Current State: has identified data integration issues and has launched several initiatives to standardize a minimum set of data definitions that will leverage the benefits of data integration within their different systems. Example: Venezuela 11Benefits from Integrated, Sharable Data cont.+ • Improve operational coordination across subunits (between units) – As the interdependence between subunits