Corporate Release 2017 R1 Development & Enhancement Repository Published by the International Software Benchmarking Standards Group 05-2017
1 Table of Contents Table of Contents... 1 Introduction... 2 Executive summary... 2... 4 Project origin... 4 Project context... 6 Type of project... 8 Type of product... 10 Product size... 10 Application group... 15 Application type... 16 Architecture... 17 Development environment... 18 Methodologies and Techniques... 22 Appendix 1 - ISBSG... 25
2 Introduction This document provides details of the various project data types that are included in the ISBSG D&E (Development & Enhancement) repository, May 2017. You will note that the project totals shown at the bottom of the tables rarely equal the 8,012 projects in the Repository. This is because submitters do not necessarily provide project data for all the data fields that ISBSG offers. The ISBSG Field Descriptions May 2017 document explains the contents of the various data fields that the Repository caters for. By studying the demographics that follow, you will be able to establish the areas that are of specific interest to you. The data in the Repository projects have come from over thirty countries. This is what makes the ISBSG Repository unique. A broad range of project types from many industries and many business areas are available for you to use for estimating, awareness of trends, comparison of platforms and languages or benchmarking. Executive summary The projects in the Repository cover a broad cross-section of the software industry. In general, they have a business focus. Project origin The projects have been submitted from 26 different countries. Major contributors are the United States (27,4% of all projects), Spain (12,7%), Australia (10,9%), Japan (10,8%), Netherlands (9,3%), Finland (7,7%), France (5,9%), India (4,2%), Canada (4,0%) and Denmark (2,2%). The projects were performed in 32 different countries. Major contributors are Spain (18,0% of all projects where the country of effort is known), United States (14,5%), Netherlands (13,0%), Finland (10,8%), France (8,5%), Australia (7,4%), India (6,9%), Japan (5,0%), Canada (3,7%) and Denmark (3,2%). Project context Industry sector: major sectors are communications (22,9% of all projects where the organization type is known), insurance (17,8%), manufacturing (12,4%), government (11,6%), banking (8,3%), medical and health care (7,7%), financial (6,0%) electronics/computers (2,9%) and service industry (2,8%). Business area: major areas are telecommunications (21,6% of all projects where the business area is known), insurance (12,1%), banking (8,3%), transport/logistics (7,4%), finance (6,9%), manufacturing (6,4%), and sales and marketing (4,3%). Type of project Development type: 67,9% are enhancement projects, 30,7% are new developments, and 1,3% are re-developments. Intended market: 84,1% of projects are developed for internal use, (i.e. for the organization that contributed the project to the Repository), and 15,3% for external use. 40,7% are developed in-house and 58,8% are outsourced. In total, 49,8% are developed in-house for internal use.
3 Team size: 29,7% of projects have up to 4 people in the development team, 30,3% have 5 to 9 people, 17,4% have 10 to 19 people, and 22,6% have 20 or more people. Type of product Product size: while IFPUG projects dominate the Repository, COSMIC, NESMA and FiSMA are all well represented. Among the IFPUG projects, 42,6% of projects have fewer than 100 UFPs, 33,4% have 100 299 UFPs, 18,3% have 300 999 UFPs and 5,8% have more than 1000 UFPs. Application group: 89,8% are business applications, 5,1% are real-time applications, and 4,2% are mathematically-intensive applications. Architecture: 35,5% of projects for which this information is available have a clientserver architecture, and 15,1% have a multi-tier architecture (there is some overlap between these groups of projects). 36,3% are stand-alone systems. Development environment Platform: 33,4% are mainframe projects, 11,5% midrange, and 19,3% personal computers. 35,9% of projects involve multiple platforms. Language: over 100 programming languages are represented. 3GLs represent 63,9% of projects, 4GLs 32,9%, and application generators 2,7%. Major languages are COBOL, C, C++, C#, Java, JavaScript, Visual Basic, PL/I, Oracle,.Net, SQL, Natural and ABAP. Development methods For ISBSG purposes a methodology applies to the whole project development process. This is distinct from techniques, which apply to individual activities within the development process. Methodology: 76,1% of projects that describe methodologies report using a waterfall model. Other methodologies include Joint Application Development (5,6%), Agile and/or RUP (5,3%), Rapid Application Development (4,5%), Multifunctional teams (4,0%) and Timeboxing (2,7%).
4 Project origin Country of origin Projects have been contributed from 26 different countries. Country of origin United States Spain Australia Japan Netherlands Finland France India Canada Denmark Brazil United Kingdom China Other 0 500 1000 1500 2000 2500 Figure 1 - country of origin Country of origin Projects Percentage Countries N % United States 2140 27,4% Spain 993 12,7% Australia 847 10,9% Japan 841 10,8% Netherlands 723 9,3% Finland 599 7,7% France 458 5,9% India 324 4,2% Canada 315 4,0% Denmark 172 2,2% Brazil 154 2,0% United Kingdom 89 1,1% China 66 0,8% Other 83 1,1% Total 7804 100% Table 1 - country of origin
5 Country of effort Country of effort Spain United States Netherlands Finland France Australia India Japan Canada Denmark Brazil United Kingdom China Other 0 200 400 600 800 1000 1200 Figure 2 - country of effort Country of effort Projects Percentage Countries N % Spain 987 18,0% United States 796 14,5% Netherlands 710 13,0% Finland 592 10,8% France 463 8,5% Australia 405 7,4% India 378 6,9% Japan 275 5,0% Canada 205 3,7% Denmark 173 3,2% Brazil 154 2,8% United Kingdom 113 2,1% China 66 1,2% Other 162 3,0% Total 5479 100% Table 2 - country of effort
6 Project context Industry sector The Industry Sector summarizes the industry, or type of organization, for which each project has been developed. Industry sector Communication Insurance Manufacturing Government Banking Medical & health care Financial Electronics / computers Service industry Wholesale / retail Utilities Professional services Other 0 200 400 600 800 1000 1200 1400 1600 1800 Figure 3 - industry sector Industry sector Projects Percentage Industries N % Communication 1533 22,9% Insurance 1189 17,8% Manufacturing 831 12,4% Government 773 11,6% Banking 558 8,3% Medical & health care 513 7,7% Financial 402 6,0% Electronics / computers 192 2,9% Service industry 190 2,8% Wholesale / retail 135 2,0% Utilities 79 1,2% Professional services 63 0,9% Other 233 3,5% Total 6691 100% Table 3 - industry sector
7 Business area This is the business area within the organization/industry that the project/application will be supporting. Business area Communications, networks Insurance Banking Transport, logistics Government, public administration Financial Manufacturing Sales, marketing Engineering Accounting Customer relations Personnel Inventory, purchasing Other 0 50 100 150 200 250 300 350 400 450 500 Figure 4 - business area Business area Projects Percentage Areas N % Communications, networks 436 21,6% Insurance 244 12,1% Banking 167 8,3% Transport, logistics 149 7,4% Government, public administration 143 7,1% Financial 139 6,9% Manufacturing 130 6,4% Sales, marketing 87 4,3% Engineering 60 3,0% Accounting 55 2,7% Customer relations 48 2,4% Personnel 41 2,0% Inventory, purchasing 30 1,5% Other 289 14,3% Total 2018 100% Table 4 - business area
8 Type of project Development type A detailed explanation of the development types is given in Appendices, Glossary of Terms. Development type Enhancement New development Re-development Other 0 1000 2000 3000 4000 5000 6000 Figure 5 - development type Development type Projects Percentage Types N % Enhancement 5426 67,9% New development 2454 30,7% Re-development 102 1,3% Other 15 0,2% Total 7997 100% Table 5 - development type Intended market This defines the relationship between the customer, the project/application developer, and application user. If the customer and the developer are in the same organization, the project is assumed to be an in-house development; if the customer and user are in the same organization the project is assumed to be developed for internal use. For some projects, it is possible to determine whether the development was in-house or outsourced, or whether the users are internal or external, but not both.
9 Intended market Outsourced for internal business unit Developed in-house for internal business unit Outsourced for external business unit Developed in-house for external business unit Multi-supplier project 0 200 400 600 800 1000 1200 1400 Figure 6 - intended market Intended market Projects Percentage Markets N % Outsourced for internal business unit 1298 49,8% Developed in-house for internal business unit 892 34,3% Outsourced for external business unit 232 8,9% Developed in-house for external business unit 167 6,4% Multi-supplier project 15 0,6% Total 2604 100% Table 6 intended market Team size This is the maximum number of people in the development team at any given time in the project. Team size 1 or 2 3 or 4 5 to 9 10 to 14 15 to 19 20 to 40 41 to 100 101+ 0 100 200 300 400 500 600 700 800 Figure 7 - team size
10 Team size Projects Percentage Catergories N % 1 or 2 331 13,7% 3 or 4 388 16,0% 5 to 9 732 30,3% 10 to 14 282 11,7% 15 to 19 138 5,7% 20 to 40 297 12,3% 41 to 100 196 8,1% 101+ 54 2,2% Total 2418 100% Table 7 - team size Type of product Product size Size is measured in function points. The 4 main function point counting approaches represented in the Repository are IFPUG CPM 4.0 or later, COSMIC, FiSMA and NESMA. Other approaches represented in the Repository include Mark II, Feature Points, and older versions of IFPUG (IFPUG 2, IFPUG 3) but there are few such projects and very few have been contributed to the Repository for many years now. IFPUG 4+ projects dominate the Repository. The numbers of COSMIC, FiSMA and NESMA projects are steadily increasing. The following tables and histograms show the range of project sizes, for each of these 4 function point counting approaches.
11 IFPUG 4+ The table shows the sizes (in UFPs) of projects sized with IFPUG function points, that are known or presumed to have been sized using CPM4.0 or later. Product size - IFPUG 4+ 1800 1600 1400 1200 1000 800 600 400 200 0 0 to 9 (XXS) 10 to 29 (XS) 30 to 99 (S) 100 to 299 (M1) 300 to 999 (M2) 1000 to 2999 (L) 3000 to 8999 (XL) 9000 to 17999 (XXL) 18000 or more (XXXL) Figure 8 - product size IFPUG 4+ Product size IFPUG 4+ Projects Percentage Catergories N % 0 to 9 (XXS) 171 8,5% 10 to 29 (XS) 363 18,0% 30 to 99 (S) 1511 74,9% 100 to 299 (M1) 1604 79,5% 300 to 999 (M2) 878 43,5% 1000 to 2999 (L) 250 12,4% 3000 to 8999 (XL) 25 1,2% 9000 to 17999 (XXL) 3 0,1% 18000 or more (XXXL) 0 0,0% Total 4805 100% Table 8 - product size IFPUG 4+
12 COSMIC 160 Product size - COSMIC 140 120 100 80 60 40 20 0 0 to 9 (XXS) 10 to 29 (XS) 30 to 99 (S) 100 to 299 (M1) 300 to 999 (M2) 1000 to 2999 (L) 3000 to 8999 (XL) 9000 to 17999 (XXL) 18000 or more (XXXL) Figure 9 - product size COSMIC Product size COSMIC Projects Percentage Catergories N % 0 to 9 (XXS) 32 7,1% 10 to 29 (XS) 45 9,9% 30 to 99 (S) 139 30,7% 100 to 299 (M1) 137 30,2% 300 to 999 (M2) 78 17,2% 1000 to 2999 (L) 22 4,9% 3000 to 8999 (XL) 0 0,0% 9000 to 17999 (XXL) 0 0,0% 18000 or more (XXXL) 0 0,0% Total 453 100% Table 9 - product size COSMIC
13 NESMA 250 Product size - NESMA 200 150 100 50 0 0 to 9 (XXS) 10 to 29 (XS) 30 to 99 (S) 100 to 299 (M1) 300 to 999 (M2) 1000 to 2999 (L) 3000 to 8999 (XL) 9000 to 17999 (XXL) 18000 or more (XXXL) Figure 10 - product size NESMA Product size NESMA Projects Percentage Catergories N % 0 to 9 (XXS) 19 3,7% 10 to 29 (XS) 35 6,8% 30 to 99 (S) 126 24,4% 100 to 299 (M1) 193 37,3% 300 to 999 (M2) 120 23,2% 1000 to 2999 (L) 22 4,3% 3000 to 8999 (XL) 2 0,4% 9000 to 17999 (XXL) 0 0,0% 18000 or more (XXXL) 0 0,0% Total 517 100% Table 10 - product size NESMA
14 FiSMA 250 Product size - FiSMA 200 150 100 50 0 0 to 9 (XXS) 10 to 29 (XS) 30 to 99 (S) 100 to 299 (M1) 300 to 999 (M2) 1000 to 2999 (L) 3000 to 8999 (XL) 9000 to 17999 (XXL) 18000 or more (XXXL) Figure 11 - product size FiSMA Product size FiSMA Projects Percentage Catergories N % 0 to 9 (XXS) 4 0,8% 10 to 29 (XS) 17 3,2% 30 to 99 (S) 99 18,8% 100 to 299 (M1) 203 38,4% 300 to 999 (M2) 165 31,3% 1000 to 2999 (L) 36 6,8% 3000 to 8999 (XL) 3 0,6% 9000 to 17999 (XXL) 1 0,2% 18000 or more (XXXL) 0 0,0% Total 528 100% Table 11 - product size FiSMA
15 Application group The application type identifies the type of application being addressed by the project (e.g. information system, transaction/production system, process control.) As there are hundreds of different application types recorded, they are grouped here into 4 groups. Application group Business application Real-time application Mathematically-intensive application Infrastructure software 0 1000 2000 3000 4000 5000 6000 Figure 12 - application group Application group Projects Percentage Categories N % Business application 4867 89,8% Real-time application 274 5,1% Mathematically-intensive application 227 4,2% Infrastructure software 51 0,9% Total 5419 100% Table 12 - application group
16 Application type A finer-grained breakdown of application types follows. Application type Financial Transaction Process/Accounting Transaction/Production System Management Information System Customer management Process control, sensor control, real time Web, E-Business Database, Catalogue/register of events or things Financial Sales & Marketing Billing Workflow management Office information system, Executive information Inventory / Ordering Analysis & Reporting Network Management, Communications Electronic Data Interchance Document management Logistics Software tool / Utility Project management Data warehouse Other 0 200 400 600 800 1000 1200 1400 Figure 13 - application type
17 Application type Projects Percentage Categories N % Financial Transaction Process/Accounting 1182 23,4% Transaction/Production System 510 10,1% Management Information System 409 8,1% Customer management 284 5,6% Process control, sensor control, real time 232 4,6% Web, E-Business 181 3,6% Database, Catalogue/register of events or things 161 3,2% Financial 146 2,9% Sales & Marketing 132 2,6% Billing 129 2,6% Workflow management 122 2,4% Office information system, Executive information syste 121 2,4% Inventory / Ordering 121 2,4% Analysis & Reporting 107 2,1% Network Management, Communications 89 1,8% Electronic Data Interchance 74 1,5% Document management 62 1,2% Logistics 54 1,1% Software tool / Utility 50 1,0% Project management 39 0,8% Data warehouse 15 0,3% Other 834 16,5% Total 5054 100% Table 13 - application type Architecture Two broad types of system architecture are represented in the Repository: client-server (of various flavours), and multi-tier (of various flavours). Stand-alone systems are also recorded as a contrast to client-server systems. Architecture Stand alone Client server Multi-tier Multi-tier client server Multi-tier with web interface 0 200 400 600 800 1000 1200 1400 1600 Figure 14 - architecture
18 Architecture Projects Percentage Categories N % Stand alone 1403 36,3% Client server 1372 35,5% Multi-tier 585 15,1% Multi-tier client server 276 7,1% Multi-tier with web interface 230 5,9% Total 3866 100% Table 14 - architecture Development environment Development platform Development platform Multi Mainframe Personal computer Mid-range 0 500 1000 1500 2000 2500 Figure 15 - development platform Development platform Projects Percentage Categories N % Multi 1986 35,9% Mainframe 1849 33,4% Personal computer 1068 19,3% Mid-range 636 11,5% Total 5539 100% Table 15 - development platform
19 Type of programming language There are many languages recorded in the repository. This can make it difficult to compare some projects. Consequently, languages are classified by type as shown below. Type of programming language 2nd generation language 3rd generation language 4th generation language 5th generation language Application generator 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Figure 16 - type of programming language Type of programming language Projects Percentage Categories N % 2nd generation language 22 0,3% 3rd generation language 4181 63,9% 4th generation language 2152 32,9% 5th generation language 13 0,2% Application generator 176 2,7% Total 6544 100% Table 16 - type of programming language Over 120 programming languages are represented in the Repository. 3 rd generation languages dominate, but 4 th generation languages are also very well represented. Some languages (e.g. Visual Basic, Visual C++) were nominated sometimes as 3GLs and sometimes as 4GLs. The table above tallies the language types as originally nominated. In the following tables, each language is consolidated under a single type.
20 Primary programming languages 3rd generation languages This is the programming language that has been nominated by the project submitter as the primary programming language. Primary programming languages - 3rd GLs Java COBOL Visual Basic PL/I C++ C C# PL/SQL Scripting Other 0 200 400 600 800 1000 1200 1400 Figure 17 - primary programming languages 3rd generation languages Primary programming languages Projects Percentage 3rd generation languages N % Java 1268 29,5% COBOL 1020 23,7% Visual Basic 440 10,2% PL/I 401 9,3% C++ 313 7,3% C 311 7,2% C# 135 3,1% PL/SQL 123 2,9% Scripting 91 2,1% Other 202 4,7% Total 4304 100% Table 17 - primary programming languages 3rd generation languages Other 3rd generation languages in the Repository include JavaScript, Smalltalk, HTML, Ada, Pascal, Periphonics and FORTRAN.
21 Primary programming languages 4th generation languages Primary programming languages - 4th GLs ORACLE.Net SQL ABAP NATURAL Delphi ASP Access PowerBuilder Lotus Notes Other 0 50 100 150 200 250 300 350 Figure 18 - primary programming languages 4th generation languages Primary programming languages Projects Percentage 4th generation languages N % ORACLE 306 22,2%.Net 277 20,1% SQL 150 10,9% ABAP 131 9,5% NATURAL 103 7,5% Delphi 67 4,9% ASP 66 4,8% Access 56 4,1% PowerBuilder 47 3,4% Lotus Notes 30 2,2% Other 143 10,4% Total 1376 100% Table 18 - primary programming languages 4th generation languages Other 4GLs represented in the Repository include Easytrieve, CLIPPER, ColdFusion, Ingres, FOCUS, IDEAL, and RALLY.
22 Application generators Application generators COOL:GEN TELON HPS Other 0 20 40 60 80 100 120 Figure 19 - application generators Application generators Projects Percentage Programming languages N % COOL:GEN 109 68,6% TELON 34 21,4% HPS 14 8,8% Other 2 1,3% Total 159 100% Table 19 - application generators Few projects that used application generators have been contributed to the Repository in recent years. The most recent projects that used application generators were implemented in 2008. Methodologies and Techniques These describe the various methodologies and techniques that may have been used during the execution of a project. They have not been related to specific project activities, and therefore may apply to any part of the development lifecycle. For ISBSG purposes a methodology (Agile, JAD, Waterfall etc.) applies to the whole project development process. This is distinct from techniques (Data Modelling, OO Analysis etc.), which apply to individual activities within the development process. Some projects mention more than one methodology (e.g. some JAD projects also use RAD and/or timeboxing), and some mention more than one technique.
23 Methodology Methodology Waterfall Joint Application Development (JAD) Agile, RUP Rapid Application Development (RAD) Multi-functional teams Timeboxing Other DevOps 0 500 1000 1500 2000 2500 3000 Figure 20 - methodology Methodology Projects Percentage Methodolies N % Waterfall 2608 76,1% Joint Application Development (JAD) 191 23,3% Agile, RUP 181 28,8% Rapid Application Development (RAD) 154 34,4% Multi-functional teams 138 46,9% Timeboxing 94 60,3% Other 57 91,9% DevOps 5 100,0% One or more of the above 2802 Table 20 - methodology Of the 191 JAD projects, 37 also mention RAD, 45 also mention multi-functional teams, and 13 also mention timeboxing. Other methodologies include Spiral and Incremental.
24 Specification, design and development techniques The following graph and table combine information from all three of these fields, as well as considering specification documents, and design documents. Specification, design and development techniques Data modeling Object Oriented Analysis/Design, UML Prototyping Business area modeling Standards (ISO 9000; CMM, CMMI) Reviews, inspections, walkthroughs Process modeling Regression testing Stress / Load / Performance testing Event modeling Pair testing 0 200 400 600 800 1000 1200 1400 1600 1800 Figure 21 - specification, design and development Specification, design and development techniques Projects Percentage Techniques N % Data modeling 1549 21,6% Object Oriented Analysis/Design, UML 1436 20,1% Prototyping 1218 17,0% Business area modeling 901 12,6% Standards (ISO 9000; CMM, CMMI) 524 7,3% Reviews, inspections, walkthroughs 521 7,3% Process modeling 485 6,8% Regression testing 225 3,1% Stress / Load / Performance testing 161 2,2% Event modeling 100 1,4% Pair testing 38 0,5% One or more of the above 3071 Table 21 - specification, design and development
25 Appendix 1 - ISBSG The ISBSG is a not-for-profit organization founded in 1997 by a group of national software metrics associations. Their aim was to promote the use of IT industry data to improve software processes and products. ISBSG is an independent international organization that collects and provides industry data of software development projects and maintenance & support activities in order to help all organizations (commercial and government, suppliers and customers) in the software industry to understand and to improve their performance. ISBSG sets the standards of software data collection, software data analysis and software project benchmarking processes and is considered to be the international thought leader in these practices. The ISBSG mission is to help YOU and your organization improve the estimation, planning, control and management of your IT software projects and/or maintenance and support contracts. To achieve this: ISBSG maintains and grows 2 repositories of IT software development/maintenance & support data. This data originates from trusted, international IT organizations and can be obtained for a modest fee from the website /project-data/ Help us to collect data ISBSG is always looking for new data. In return for your data submission, you receive a free benchmark report that shows the performance in your project or contract against relevant industry peers. Please submit your data through one of the forms listed on http://isbsg.org/submit-data/ Partners This page will help you to find an ISBSG partner in your country http://isbsg.org/meet-isbsgpartners/