The first generation of organizations that have tried to ‘manage' knowledge have found that people could not be compelled to use the new systems - and, without constant updating, the systems languished. Focusing attention on usability, convenience and people's experience in using a system is more likely to increase take-up.
Flexibility at the front-end has to be balanced with clarity and single-mindedness behind the scenes.
Consultants can provide help in coordinating stakeholders and suppliers, as well as expert help in terms of the technology selected, but their key role is to provide a bridge between the front and back office.
Helping 300,000 citizens pilot new voting channels required careful attention to communications and tight control of project tasks.
Voting procedures in the UK have not fundamentally changed in the last 100 years. During that time the UK has resisted every mechanical form of voting machine, and while the nation has avoided any controversy caused by ‘hanging chads' or other troublesome artefacts of voting machinery, it has paid the price in expensive, labour-intensive and time-consuming election processes. While the world at large has struggled to eliminate paper from its daily business, democracy in the UK still relies on individuals using pencils to make crosses on ballot papers at local polling stations. Selected schools, community centres and other local amenities are closed to normal business during each polling day. Fleets of vans carry metal voting boxes to centralized venues for counting by hand. If someone were paying for this unwieldy and outmoded process, it would be reformed with all due speed. And of course someone is paying - the taxpayer. Although the current e-election pilots are not designed to demonstrate value for money, this will be a major consideration in the government's future modernization programme.
The government had made explicit commitments to implement electronic service delivery across all public services, setting a target of 100 per cent availability of services electronically by 2005. As part of this drive for modernization of the relationship between the citizen and government, voting could not be left behind. The government commissioned e-election pilots to test UK citizens' appetite to vote by whatever means reflects their lifestyle. E-voting would therefore be a very visible plank of e-government. Unisys's Phil Cheetham sees e-voting as an inevitable development: ‘Citizens can bank online and shop online, and all other government transactions are being enabled online, so the voting system shouldn't be left behind.'
A government-backed report published by De Montford University in May 2002 confirmed emerging public support for the idea of Internet and telephone voting, but warned that the government would have to help educate those electors who were less confident with new technology if an electronic General Election could go ahead in the future. The report also endorsed the government's policy of trialling e-voting in local elections, through ‘multi-channel' voting, which offers voters a range of options including the traditional polling station and ballot box. Electronic voting had been used for the first time in the 2000 local council elections, and helped lift voter turnout in some areas. However, postal voting had proved the most popular voting method, almost doubling turnout in some places where it was piloted.
The government committed to extending the pilots, with a view to considering an e-enabled general election after 2006. The local government minister, Nick Raynsford, said that the government was keen to make the electoral process ‘more relevant to modern life', but warned that ‘any changes to the voting system must be properly researched to ensure that they are of real benefit to the public, as well as incorporating effective safeguards against abuse'.
The De Montford report had in fact raised public concerns over secrecy and security, with possible threats from viruses, power supply disruption, hacking and limits to system capacity. People feared that the risks inherent in large commercial online systems could transfer to the electoral arena. Technology could be used to improve the voting process, but maintaining public confidence in the integrity of the process needed to be the highest priority in any new approach.
The job of managing the UK's Electoral Modernization Programme is funded by grants to a number of local authorities and managed by the Office of the Deputy Prime Minister (ODPM). The ODPM was created in May 2002, with responsibility for policy on housing, planning, devolution, regional and local government and the national fire service. It also takes responsibility for the Social Exclusion Unit, the Neighbourhood Renewal Unit and the Government Offices for the Regions.
Unisys and its partner, Election Systems & Software (ES&S), were appointed in March 2003 to deliver multi-channel voting systems to the citizens of South Tyneside, Chorley and Rushmoor Councils in the May 2003 local elections. The combined electorate of the three areas was more than 300,000 people. Unisys and ES&S were also contracted to provided data hosting services for a further five local authorities whose ‘end-user' election services were managed by BT.
The pilot options
Multi-channel voting covers a number of different methods for registering votes, not all of them high-tech. In an all-postal ballot, for example, the local authority posts ballot papers to all those registered to vote. Usually included with the ballot paper is a ‘Declaration of Identity', which the voter and, in some cases a witness, are required to sign.
‘Early voting' is the name given to an extension of normal voting hours. Confusingly, early voting can also include polling stations closing later than usual, or even remaining open for additional days. Where polls are open for additional days, councils can use mobile polling stations, located at convenient sites such as supermarkets and train stations. Early voting is regarded as a novel voting channel, even though it is only a procedural variation on the traditional voting arrangement.
However, electronic methods also have a part to play in the improvement of traditional voting methods. ‘E-counting' is a means of automating the counting of paper ballots either by scanning the paper directly into a computer as an image or printing bar codes on the ballot paper next to each candidate's name for scanning.
In Internet voting, electors are each supplied with a unique security code. They log on to the voting site and enter their code. The system then checks their eligibility to vote and displays the voting information for the relevant electoral ward. Telephone voting uses automated response systems, familiar from commercial helpline services, to register votes while the SMS or text messaging option exploits the popularity of this method of communicating from mobile phones. Voters are sent a personal identification number (PIN) and a pack listing the names of the candidates standing in their ward and their corresponding candidate numbers. Voters then send a text message, including their PIN and the identifier of the candidate they wish to vote for, to the telephone number supplied with the voting pack.
Piloting the alternative channels
The three pilots each used a different mix of technologies to deliver new voting choices. In South Tyneside electors were able to vote via post, the Internet, touch-tone telephone, as well as via SMS text messaging and multilingual touch-screen kiosks in six locations around the city. Postal votes were counted electronically rather than manually. The voting period was from 15 April until 5.00 pm on 1 May with the exception of the SMS option, which closed 24 hours earlier.
Chorley citizens were able to vote by post with electronic counting, by Internet and touch-tone telephone from 18 April until 9.00 pm on 1 May. Citizens in Rushmoor combined voting at traditional polling stations on 1 May with Internet voting from 25 April until 9.00 pm on 1 May.
The e-voting pilot schemes had four objectives. The first objective was to make voting more straightforward for the public by offering simple methods for registering their votes. The second objective was to make elections more accessible, by making it more convenient to vote and by making voting more attractive to people currently less likely to vote. The third objective addressed the behind-the-scenes aspects of elections, by making the administration of elections more efficient and cost-effective. The final objective of the schemes was to maintain or even increase the level of security at elections.
The Unisys project team also added a management objective to collect and analyse operational information in order to assess the scalability of each method for subsequent pilots and potentially a General Election. The focus of the data analysis was to help future elections to be run more efficiently and resources to be deployed more effectively. The team also wanted to discover how long it took for the local authorities to become familiar with the technologies used in the pilots. This information could then be fed into the planning and milestones for future e-voting pilots.
The government required every supplier in the e-voting pilots to demonstrate innovation. The team were therefore given considerable freedom in how they managed and delivered each pilot. The pilots were set in areas with diverse demographic characteristics. The project was further complicated by the fact that each technical solution was different and the voting periods varied across the areas.
The three pilots were implemented using Unisys TEAMmethod™ project management methodology. TEAMmethod™ is a formal approach to designing, implementing and maintaining information systems organized in four phases: Strategy and vision, Planning and road map, Deployment and review, and Validation. Several components of this flexible methodology were implemented including acceptance, quality and testing. A primary focus was placed on strong management of risks and issues.
The centralized ES&S software application was installed at a secure Unisys data centre in Milton Keynes and contained records of all registered voters and their voting activity. The application was based on an Oracle database and Oracle 9iAS server. All the source code was reviewed by an independent testing authority to ensure that it complied fully with European coding standards. Centralization of the e-voting systems meant that there was no technical disruption to the systems environments of the three authorities.
A team of experienced project managers was placed with the local authorities to work in delivering the elections. A central management function and project office managed the team. A centre of excellence including skills for election delivery, the solution and its technology and processes was formed to support the local authority pilot project management. A round-the-clock helpdesk, with communication channels through to the local authorities and the centre of excellence, was put in place to address any queries from voters.
Dealing with the issues
Expanding the voting options available to electors is not simply a matter of installing the right technologies. Indeed, the technologies used for the pilots were all proven solutions that had been used in a range of commercial environments. The public were not going to be guinea pigs for new technology solutions. The team's concern was more focused on how the alternative voting channels would be delivered, so that the technologies could yield their expected benefits and the performance of the pilots could be adequately assessed. Changing electoral methods is about influencing behavioural changes and determining which mechanisms have the greatest impact on turnout and cost-effectiveness.
The first issue the team faced was the various local authorities' familiarity with the technologies being applied in the pilot. Some officers were concerned about the consequences for public confidence if a channel failed during the ‘live' voting period. The twin spotlights of government scrutiny and public opinion were firmly on the pilots and any glitches would be magnified in media coverage. It was vital for the team to keep the authorities constantly up to date with project developments and to take them through each stage of the process at an appropriate pace. The authorities needed to feel that they were in control of the pilots, and that the technology tail was not wagging the democracy dog.
Having chosen the channels to implement, the local authorities needed to agree the voting processes and the look and feel for each channel. For Rushmoor, where the Internet was the only novel channel being provided, designing the voter experience was a relatively simple process. The team created a set of templates for the process and walked colleagues from the authority through the screens. Each screen was then laid out to meet local preferences, with the content and style of each screen documented, implemented and agreed. At South Tyneside there were five channels to implement including e-counting, Internet and SMS text messaging. The designs of all these channels were completed and agreed in the same open and collaborative style.
With the channels correctly tailored for each voting area, the team then turned to the task of ensuring that each local authority team had enough support to deliver the pilots. A central support team was rapidly formed and briefed on the project's goals, timescales and predicted challenges. This team developed into a lively, supportive community linking the different pilot sites and disseminating knowledge among them. The cross-fertilization between the different pilot areas meant that potential obstacles were identified across the community as soon as they arose in one area. Solutions could also be conveyed to the wider team before the problems were even encountered. In this way a virtuous cycle of learning and improvement took root during the pilot exercises themselves. This inductive learning was an additional bonus to the knowledge the team expected to gain at the completion of the pilots, when overall performance could be compared to traditional methods.
The third issue the team faced was voter awareness of the pilot channels. The team developed a ‘Voter Outreach' campaign to increase the public's awareness of the new technologies being made available for the upcoming elections.
Wider awareness campaigns were run in each area to reinforce the personal communications packs and encourage voter engagement with the pilots. In Chorley, the team worked with Chorley Borough Council to devise and deploy a ‘voting at your fingertips' marketing campaign. This campaign targeted several communications channels simultaneously in order to ensure complete coverage of the local area in the key period of the run-up to the elections. The team ran a local press campaign to make sure that relevant stories were published regularly before and during the election period. The team also designed an advertisement to run on the main local radio station throughout the 12-day voting period. Posters using the distinctive ‘voting at your fingertips' logo were placed at a selection of prominent roadside billboard sites, and posters highlighting the voting channels, voting period and helpline number were displayed in community sites such as doctors' surgeries and post offices. Finally the team ran an open day on the middle Saturday of the voting period. Sited in Chorley town centre, the open day event allowed voters to ask questions face to face with team members and acted as a lively reminder of their voting options to passers-by.
Central to each pilot was a ‘voting pack' containing all the information about the channels available in each voting area. Every elector within each constituency received one of the packs, ensuring that no one would be excluded from the alternative channels by accident or lack of access to general information sources. The pack was practical as well as informative, containing simple step-by-step instructions for voting by each voting method. A helpline was also set up so that any citizens with questions about voting channels or procedures could call one central number and have their queries answered personally and in full.
The authorities were highly aware of their leadership position in trialling the new voting methods. Chorley's Deputy Returning Officer Martin O'Loughlin says: ‘By increasing the choice of Internet and touch-tone we are making voting even easier. Britain is ahead of most countries in moving to e-voting and Chorley is leading the way nationally and now internationally.'
The team identified and tackled all these issues within the context of a fundamental constraint: the project's timeline. The timetable for delivery was short and non-negotiable. For the delivery of a voting system to be operational in mid-April when it was ordered in mid-March, the project had to be tightly managed so that there were no delays in any stages of the project. Regular team briefings meant that issues were overcome almost immediately. The immovable timeline was a helpful constraint on the project, guarding against any potential for ‘scope creep' or the addition of extra options. According to Cheetham: ‘There was no time to dwell on problems, we had to fix them.'
The Unisys consultants on site at each local authority were integrated as part of the election project teams and participated in regular project meetings and status reporting. Where specific areas of expertise were required, the appropriate representatives were called in for those meetings.
Reporting to the central management structure spanning all the pilots was done by weekly status reports. These reports included risk and issue logs. Weekly meetings were attended by at least one person from each local authority site. A teleconference link was made available for anyone who could not attend in person, but this was rarely used since most team members felt the value of personal attendance.
Unisys also worked in close partnership with the ODPM throughout the election period and afterwards, regularly discussing and advising on issues relating to service levels, detection of attempted security breaches and the progress of the various public awareness campaigns. The team also worked collaboratively with other government agencies including the Electoral Commission. The team's practice was to maintain an ‘open door' policy for all stakeholders in the project, inviting attendance at internal project meetings as well as site visits to our project sites.
Over to the count
When the voting is over, the counting begins and the shape of the public's democratic decision emerges. With the end of the elections the project team was able to measure the uptake and performance of each pilot channel to see how voters had expressed their preferences for the way they vote.
Analysis by the Electoral Commission showed that the overall level of user awareness and comprehension of the voting methods being tested were high, with 77 per cent of those interviewed during the election and 88 per cent interviewed afterwards confirming that they were aware of the new voting arrangements. Analysis also showed that voters were appreciative of the alternative channels. Independent MORI polls reported that respondents felt that the convenience of e-voting channels was on average ‘very good' (40 per cent) or ‘fairly good' (27 per cent). Those who said that the channels were neither better nor worse than traditional channels totalled 8 per cent, 7 per cent said the new channels were ‘fairly poor', 3 per cent ‘very poor' and 14 per cent had no opinion. Opinion polling also revealed local variations in voter responses, with three-quarters of respondents to the Electoral Commission's MORI poll in Chorley saying that the new voting methods made the process of voting better.
On the technology side of the project, independent security tests carried out on the electronic channels failed to penetrate the security of any channel. The team could also find no evidence that the alternative voting procedures led to any increase in electoral offences or malpractice.
The team refuses to mistake ‘soft' benefits for measurable effects, but it does point to several outcomes it believes make a significant contribution to the progress of the electoral system. For instance, voting turnout is influenced by a number of factors and politicians frequently disagree about why numbers go up or down at different elections and in different areas. The impact of e-voting on turnout is particularly hard to assess as there is so little historical data. However, as with the 2002 local elections, the turnout for Chorley in 2003 (49.81 per cent) was substantially greater than historic turnout figures for traditionally run elections in the borough (32 per cent). Meanwhile, South Tyneside's Council Leader Paul Waggott says: ‘These e-voting methods have made this a very different election. The experiment has been a success as it has doubled the number of people voting than at traditional ballot box elections.'
A further ‘soft' benefit of the pilots has been the increase in local authority experience in running e-voting systems and their associated communications and support programmes. This should help increase the level of efficiency and reduce implementation costs for future e-voting pilots while reducing reliance on external agencies for help in implementation.
The team also notes that the cost profile of e-voting is not yet mature enough to judge in terms of standalone value. As in any new technology-dependent process, e-voting requires up-front investment that takes time to pay back. The expenditure needed to provide one-off pilots for individual authorities was borne by the ODPM. The provision of telephone and Internet voting did lead to a significant increase in expenditure over the normal cost of traditional elections, mainly for the provision of the hardware and software needed to support the new channels. Deploying this technology over a number of years and on a wider scale will provide further operational evidence to inform the assessment of the long-term cost of implementing e-elections.
There were no e-voting pilots in the 2004 local elections. The Electoral Commission advised that the simultaneous European Parliament elections would make the 2004 elections complex enough without additional voting channel trials. However, Cheetham says: ‘It's clear that e-voting works and that it should be expanded in future years.'
Unisys collaborative work with its partners in the local authorities shows that careful preparation and attention to detail can make alternative methods of voting simpler than the time-honoured practice of folding paper slips into boxes, and can help to improve the reach and effectiveness of our democratic process. The consulting firm was able to centralize systems provision for all of the pilot areas it served, saving the public from duplicated costs and avoiding disruptive work at the local authorities involved. In addition, Unisys was able to hold together the different strands of activity in each voting area, thereby providing a local focus for each authority without distracting management attention from their ongoing concerns. Detailed expertise in the various novel channels and their technical deployment would not be found in complete or up-to-date form in any of the authorities, so access to a full range of skills was important to each area's e-election.
Crucially, Unisys acted as a conduit of learning, both geographically and historically. Alerting different teams to issues already encountered reduced the numbers of errors and delays encountered across the election as a whole, while the lessons learned from the exercises will be used to improve elections in future years. Good consulting firms act a little like living libraries, helping to transport knowledge across the miles and years. This valuable experience may pay further dividends for the company and its partners as pressure mounts for publicly traded companies to move to e-voting. Phil Cheetham is convinced that e-voting will soon become the new tradition: ‘This was a challenging project because of the timescales and the complexities. But citizens and customers were delighted, and there's real zeal for continuing the pilots. The ultimate [end] will be the General e-enabled Election.'
Getting BAE System's procurement systems to talk sense with the industry's online trading environment is anticipated to make the company £20 million in cumulative savings. The flexible method the company used to streamline procurement will also spark further savings throughout the business as it is applied to a growing number of business processes.
BAE Systems employs more than 90,000 people worldwide, and generates annual sales of around £12 billion. The company designs, manufactures and supports military aircraft, surface ships, submarines, radar, communication tools, guided weapons and avionics equipment. The company operates in an economic and political climate where customer value is all-important and public scrutiny intense. This is big business: service agreements with governments and airlines are measured in the billions of pounds and product life cycles are measured in decades. It does not automatically sound like a great candidate for an e-commerce solution.
Some managers bridle when they hear the term e-commerce. The hype surrounding business-to-consumer e-commerce during the late 1990s led to some best-forgotten projects as organizations tried desperately to join the Internet revolution. While the leading e-commerce players such as Amazon and eBay have gone from strength to strength, smaller players have fared less well. Existing businesses that went online frequently did not generate the revenues they expected, and sometimes overspent on expensive Web sites that added little to the bottom line.
However, e-commerce is now silently and unobtrusively delivering significant recurring benefits to businesses around the world. These successes are in the category of business-to-business systems, and in particular applications for online procurement. Cost-effective procurement is a vital part of managing service delivery costs, so engineering paper transactions out of the process can generate substantial value.
Exostar is one such business-to-business solution. It was founded by BAE Systems together with other industry leaders Boeing, Lockheed Martin, Raytheon and Rolls-Royce as an online trading exchange for the aerospace and defence industry, linking up manufacturers, suppliers and customers. Exostar provides an online, collaborative trading system where data and component designs can be shared with suppliers, contract agreements exchanged, and parts sourced and purchased electronically, saving considerable sums of money compared to paper-based trading. Exostar's user base includes over 300 buyers in 20 different countries and more than 12,000 suppliers worldwide.
Businesses can gain instant benefits from using online procurement as a means of buying standard supplies such as stationery or furniture. Most businesses now use online sources for purchasing items that can be readily listed in catalogues. However, Exostar goes beyond the typical catalogues. Exostar members can use the system to share design documents with each other, allowing companies to collaborate on the products they are seeking to create or consume. In this way users can craft their own deals. Furthermore, once such a contract has been agreed on Exostar, the two parties can complete all their purchasing functions with each other through Exostar. In effect, Exostar provides a standardized environment in which parties can do business with each other from concept through to delivery.
Electronic trading links existed in BAE Systems' industry prior to the launch of Exostar, but these relationships carried a small proportion of transactions. Exostar created a new, open, global channel for electronic transactions that would drive much more business online. The implications for trading partners' systems were profound, not least at BAE Systems itself.
BAE Systems had inherited a host of different Enterprise Resource Planning (ERP) systems as a by-product of a series of mergers and acquisitions. Each business area had its own ERP suite used for procurement. These different procurement systems with their varying functions, standards and platforms did not form a natural electronic trading strategy for the company. The legacy base was also driving escalations in the business's operational costs. The costs of duplicated functionality and lost opportunities for efficiencies caused by the non-availability of a unified trading environment within the company were felt to be large, though hard to compute. BAE Systems estimated that 99 per cent of its purchase orders and invoices were still being processed manually. Given the scale of its global operations, and the leadership in electronic trading it had provided with the co-founding of Exostar, the company saw a massive opportunity to reduce costs and improve stakeholder value. The result is a streamlined procurement platform calculated to generate cumulative savings of £20 million by 2010.
Tasking the team
A BAE Systems/CSC Computer Sciences Corporation team was formed to enable the business to trade and exchange procurement information effectively and efficiently, both within the business and with external partners. Exostar would provide the ‘pull' for the project, while the unacceptable costs of the legacy situation provided the ‘push'. BAE Systems and CSC would work closely together throughout the project, sharing the business goals and involving each other as integral members of the design, development and deployment teams. BAE Systems' e-Business lead Andrew Mossop says: ‘The project success was all about working together as one team. [BAE Systems] and CSC worked closely together to deliver the final solution that removed non-value-added tasks from the procurement process.'
CSC's Mike Burz adds that BAE Systems' long history with CSC was an important factor in the project: ‘Over the years we'd proved to them time and again that we deliver state-of-the-art solutions that benefit them financially. And so they trusted us to come up with an innovative solution to this e-commerce problem.'
The team was asked to design, develop and deploy a solution that would enable each ERP system to connect to Exostar. The solution would include:
establishing the connections between the ERP systems and Exostar;
collaborating with BAE Systems' dedicated e-business team to develop a business case and secure buy-in from the key stakeholders within the business units;
promoting the use of the newly developed solution for other e-business applications, such as e-Finance, e-Sales, e-Sourcing and e-HR, within BAE Systems worldwide.
The team defined and evaluated three different approaches to meeting the project's goals. The first option was to create a centralized procurement system to replace the plethora of existing systems throughout the business. This would have the benefits of standardizing procurement processes within the company and removing the legacy burden. However, it would also be an expensive approach that would expose the business to critical risks. The diverse procurement processes fitted the businesses they served, and defining a single, universal replacement system might not be possible without fatally compromising some areas of the business. Building a system that offered the full flexibility needed by the company's operational units would essentially have entailed rebuilding the functionality manifested by Exostar, albeit in an in-house setting rather than a sector-wide context.
The second option was to implement a set of point-to-point solutions, where each of the existing ERP systems was connected directly to Exostar. An operational relationship would be designed and coded to give each legacy system a ‘private line' to Exostar. While this was a more viable and less risky approach than the first option, it would add to the complexity of the company's legacy systems. Each relationship would have to be hand-crafted, and there would be no opportunity for sharing data or functionality among the relationships. The team would have to build and maintain a unique bridge for each legacy system, without being able to share any building materials or components.
The third option was to create a central backbone into which all the existing ERP systems would connect. This option was dubbed the Enterprise Application Integration (EAI) approach. The EAI layer would manage two-way traffic between BAE Systems' procurement systems and Exostar, ensuring that the information was translated appropriately for each system.
Following extensive discussions, the third option was chosen because it provided the greatest benefit at low cost and minimal risk. This option was also considered to be more flexible and adaptable. One of the particular plus points was the potential to adapt the EAI strategy for other parts of the business. Not only would it be capable of unifying disparate ERP systems, it could equally be applied to enable other business processes that required streamlining, such as sales or HR, thereby supporting BAE Systems' longer-term business objectives.
A language for business
The evolution of programming solutions built around Internet technology has produced many benefits, not least the standardized user interface framework provided by the ubiquitous Web browser. The HTML (hypertext markup language) coding scheme defined for Web pages allows all kinds of information to be displayed by browsers and promises a high degree of consistency across different browser products and computer platforms. But while HTML has helped to simplify the creation of content, it remains merely a coding scheme for the visual layout of material. The items tagged in an HTML page have no business meaning. A piece of HTML may tell us, for example, that a word should be displayed in bold, but it cannot tell us that the displayed word denotes the price of the item placed alongside it.
Developers quickly realized that an analogue of HTML designed to incorporate the meaning of content as well as its styling would have a powerful effect on Web-based systems. If packages of content could also tell programs what their elements mean in business terms, then systems would be able to collaborate with each other without human intervention. Systems would be able to use common vocabularies, implemented as simple codes embedded in the data they describe, to work together.
The result is XML (extensible markup language). Developers use XML to create vocabularies that describe particular business areas, so that systems in the same business domain can communicate with each other. XML is also at the core of the development efforts of every major software vendor, ensuring a growing skills base and a broad commitment to its long-term use.
BAE Systems and CSC identified XML as the ideal solution for creating the EAI layer. Exostar provides a stable data structure that can be thought of as a target language. Each of the systems that needed to interface with Exostar spoke their own languages. But every system in this heterogeneous collection could be described thoroughly and accurately in XML. The team could then build XML adapters to pipe data between the different business languages.
A key benefit to this approach is that it has very little impact on legacy systems. As the name suggests, ‘systems integration' usually demands that the systems being integrated each be modified in order to create a new, combined entity. This approach is disruptive, and highly risky. In the first place, every legacy system running in a business such as BAE Systems is there to perform a business purpose. Taking the system offline, or replanning its maintenance schedule, in order to integrate it with another system causes interruptions in service. Even system testing following integration will present some disruption to existing users.
Developers then face a potentially greater problem associated with invasive systems integration: the risk of introducing errors to the system during redevelopment. Most legacy systems fall into the category labelled ‘if it ain't broke, don't fix it'. Legacy systems remain in organizations because they carry out tasks vital to the business. They are also usually complex systems that have been altered over many generations of development, to the point where developers may no longer be able to understand fully how they work. As layer upon layer of code builds up over the years, dependencies between parts of the system become harder to comprehend. These dependencies are often only discovered when attempts at maintenance or integration inadvertently break them, and the system begins to behave unexpectedly or erratically. Systems integration is not a task for the faint-hearted, and is understandably avoided in many parts of organizations.
Using XML adapters provides another means of getting systems to work together without tinkering with their internal workings. An XML adapter is designed to match the expected output and input of a target legacy system. It then works with that legacy system in a way that mimics what the legacy system is used to. The adapter acts rather like an antibody, attaching to the system through a communicating interface that matches the target. The opposite end of the adapter - in this case, where it ‘speaks' to Exostar - matches the online exchange's defined inputs and outputs. By building an adapter for each target system, the entire portfolio of legacy systems can be connected to Exostar without any internal adjustments, loss of business service or an extended testing cycle. The outcome is more like systems collaboration than integration. The family of systems works together through the set of adapters, which encapsulate the required business logic. In a traditional systems integration project, the code dealing with communications with Exostar would be distributed among the legacy systems. The BAE Systems/CSC strategy enables much easier management of the systems portfolio going forward, since new adapters can be readily added to the set when needs arise.
From pilot to showcase
The team was highly aware that it needed to prove the business readiness of its solution if it was to win approval from senior management. Security was a key issue. The scale and confidentiality of BAE Systems' business meant that any failure to secure every aspect of the procurement process from loss, misdirection or capture would destroy the new solution's credibility. Consequently the team began by building a proof-of-concept system limited to one site. Transactions were limited to purchase orders and purchase order amendments, and transacted initially with a small group of selected suppliers.
Customer Support and Solutions (CS&S) Operations in the UK was selected as the pilot site. CS&S is dedicated to managing service delivery costs for all customer contracts. It is a global operation with service staff based in offices all over the world. The UK operation is situated on 30 sites and typically generates 23,000 transactions per year, rising to 65,000, through no less than 32 different procurement systems.
Risk was further reduced through the use of standard, validated management methodologies. BAE Systems' LifeCycle Management methodology was used alongside CSC's Catalyst™approach to create and implement the solution, providing management control and quality assurance right the way from business case creation through to deployment and delivery to the support environment.
The team was aware that the project's success would rely on the cooperation of procurement staff both inside and outside BAE Systems. Although senior management could be bought in through demonstrable attention to issues of security and risk containment, those who use the systems everyday would be key to its delivering real benefits to the business. There was some initial resistance both from suppliers and BAE Systems staff, since the project changed the processes they were familiar with. CS&S staff were used to downloading purchase orders from their ERP system and e-mailing them to the supplier, then receiving responses to the orders by e-mail. Now they had to rely on a new system and new processes to transact orders with their suppliers.
This critical issue was overcome by running a number of workshops in the BAE Systems Solution Demonstration Facility. During the workshops both buyers and suppliers were able to try out the new system, executing a number of business case scenarios, reviewing the results and highlighting any issues that needed immediate attention.
Savings that count
The effect of connecting the first ERP system to Exostar radically changed the way the CS&S Operations department handled its procurement. The solution proved to be so successful and flexible that it was quickly deployed in several other BAE Systems business units. The roadmap for deploying the system throughout the global business is now in place.
Within CS&S Operations alone, the cost of sending orders to 6,000 suppliers was reduced by as much as £7 per transaction. This is expected to provide a return on investment of US $400,000 in the first three years, even before BAE started to calculate other benefits such as reduced cycle time and improved data quality.
The company's projected procurement savings of £20 million are impressive in themselves. However, BAE Systems has also acquired a platform for systems collaboration that will impact many crucial areas of its business. By removing the human translation of information between systems, and letting different systems share a common language via the EAI layer, BAE Systems has refitted the business for future expansion while attacking its operating cost base. The company's collaboration with CSC has also created the capability of collaboration for its many enterprise systems, and laid down a path to further, repeating bottom-line benefits.
Resolving the complex junction of a set of business applications such as BAE Systems' procurement systems is an undertaking that requires sensitive and expert handling. These kinds of systems integration projects are often best handled by external consultants, since they tend to bring the most up-to-date integration skills to the task. This is not to say that in-house staff are not equal to the task. But an external consultancy brings repeated experience of completing projects of similar scales and complexity, whereas for in-house staff an integration of this size and criticality might be a once-in-a-lifetime occurrence. The required technical skills, in this case those of XML, can be taught but their rapid application to the business environment cannot be assured. Organizations reduce their technology risk by using experienced help at such times.
CSC's position as a consulting partner also helped it perform a mediating role, ensuring that the organization's senior management were behind the project and that the right area was chosen for the pilot. The team's technical solution mediates among systems; its consulting style mediated among people.
The launch of a knowledge management portal to help regenerate the UK's most deprived neighbourhoods delivers guidance direct to users in the community.
The UK government's National Strategy for Neighbourhood Renewal aims to close the gap between deprived areas and the rest of the country within 10 years. Success depends on providing policy-makers, specialists in this field and community groups with ready answers to the question: ‘What works where?' The National Strategy Action Plan called for development of a knowledge management system, which would be: ‘a systematic and comprehensive guide to the information available on what works in tackling the various problems of deprived neighbourhoods. It will draw upon experience of what works, across England and beyond, and link into sources of evidence from other departments, outside bodies and regional, local and neighbourhood feedback.'
The Neighbourhood Renewal Unit (NRU) worked with PA Consulting Group to create a comprehensive, Web-based knowledge management portal to meet this need. The portal was developed using PA's Rapid Systems Development approach, which encouraged users to shape the portal and its content to their requirements. Despite a challenging development timescale of six months, the portal was implemented two weeks ahead of schedule.
Since its launch by Neighbourhood Renewal Minister Barbara Roche in October 2002, the site has received more than 350,000 hits. This level of traffic is well above the specified targets. Of the users surveyed 88 per cent have found the portal useful in the design and delivery of neighbourhood renewal programmes. This represents a step-change in user behaviour. As intended, neighbourhood renewal initiatives are now being based on evidence about ‘what works', accessible from a single, authoritative source.
PA's Jim Knox says:
What made this assignment so challenging was the wide range of users involved in neighbourhood renewal projects, from residents in deprived neighbourhoods to policemen, health and education professionals and policy civil servants. We had to find a common language for reaching all these groups, and helping them do their jobs better.
Inequalities in the quality of life
Over the past 20 years, hundreds of neighbourhoods throughout Britain have seen a wide gap open up between their residents' quality of life and that enjoyed by people elsewhere in the country. For example:
In the 10 per cent most deprived electoral wards, 44 per cent of people rely on means-tested benefits, compared with a national average of 22 per cent.
In the 10 per cent most deprived wards 43 per cent of all housing is not fit for habitation.
40 per cent of the population live in the 88 most deprived local authority areas.
The National Strategy for Neighbourhood Renewal aims to close the gap between deprived areas and the rest of the UK by working with policy-makers, local agencies and community members to reduce levels of crime, unemployment, ill health and under-achievement. The government set up the NRU to lead the delivery of this agenda.
The NRU recognized that access to appropriate information was essential to deliver change. However, a departmental survey showed that only 11 per cent of people involved in delivering current neighbourhood renewal programmes were basing their design and delivery on evaluated evidence, such as studies of earlier initiatives. The same survey showed that 83 per cent would use such evidence if it were available.
Based on this survey and on similar findings from the government Policy Action Team 16 (PAT), the NRU identified an immediate need to develop an online knowledge management portal for the neighbourhood renewal community, giving access to evaluated evidence of not only ‘what works' but also ‘what does not work'. For example, the portal could share experience suggesting that CCTV initiatives to improve neighbourhood security need to be supported by other measures such as community wardens or ‘alley gating'.
The NRU wanted all information to be delivered clearly and succinctly, so that it could be readily used by all kinds of people. The portal would include online networks of practitioners, policy and guidance documents, and case studies of projects that were either under way or completed.
The project called not only for skills in the design and development of knowledge systems, but also for familiarity with neighbourhood renewal activity in a large number of areas. The knowledge areas covered a complex mass of government policy, including housing, crime, employment, health and education. The team had to identify and address the key topical issues for each of these areas in order to make the site as relevant and usable as possible for its target users.
Designing the portal
The NRU/PA team made its first mark by agreeing and securing a brand for the portal: www.renewal.net. The team then set four goals - to:
clearly define the overall system requirements;
develop the www.renewal.net brand and imagery;
identify the overall look and feel of the site;
define an information model to support content classification and site navigation.
The new portal had to meet the needs of an unusually diverse user base, ranging from local residents to neighbourhood renewal practitioners. Users would therefore have varying degrees of IT literacy. PA used its rapid systems development (RSD) approach to drive out the optimal design. RSD is an iterative technique that involves users and other stakeholders in the design and development of a system from day one. RSD enabled the team to take into account the fact that renewal programmes had not historically made use of evaluated evidence. The technique ensured that users took ownership of the initiative from its inception. This was the only way to make sure that the portal included the content that users really needed, and that the users also modified their working practices to make use of that content.
The team's iterative approach also allowed it to prioritize development of the content to ensure that it matched the government's priority areas in housing and environment, health, education, worklessness and crime while designing both content and format to maximize usability of the site.
The team also had to formulate a strategy for capturing the vast amounts of content relating to neighbourhood renewal, and enabling further content to be added to the growing resource. Microsoft's SharePoint software was used as the platform for content management.
Over the six-month development life cycle, the team consulted more than 2,000 prospective users of the portal. The team held focus groups, workshops, brainstorming sessions, conferences and seminars. These consultations provided opportunities for reviewing and commenting on successive versions of the site, thereby helping to control the iterative progress of the project. The sessions also enabled the team to evaluate documents being developed for the site with the target audience.
Renewal.net's users include people from all walks of life, from senior civil servants and policy-makers to tenants within the UK's most deprived areas. Knox says that the diversity of the portal's users was a guiding factor in the team's design of the site:
There are all sorts of people involved in partnerships to renew neighbourhoods, including members of the community and staff of statutory agencies like the local authority or police. Say they've identified a problem with abandoned cars. They can go to renewal.net, search on ‘abandoned cars' and access half a dozen case studies on how the problem was tackled in other places.
The programme of consultation sessions also created visibility for the portal, and a demand for its services. The sessions allowed the team to promote the site widely prior to its official launch. On the supply side, the extensive discussions helped create a consensus on how the portal's information would be collected and displayed. This process included every government department concerned with neighbourhood renewal issues and policy.
Where suitable advice on renewal did already exist, it could be on any one of hundreds of Web sites and was therefore difficult to find. The team designed a content collection strategy that allowed users to identify and access existing policy and guidance documentation via a single contact point. In addition, templates were developed to ensure that new documents created specifically for the portal would be clear and easy to understand.
Much of the existing evidence consisted of complex, lengthy research studies. What users needed, however, was simple, punchy material written in plain English and straightforwardly answering the question ‘What has worked and why?' The team arranged for the content to be rewritten to meet this need and conducted interactive user groups to confirm that the right note was being struck.
A critical part of developing new content involved recruiting the right team of authoritative contributors across the broad range of subject areas. Such a group had never been formed before. The project team organized coaching for these independent experts in writing effectively for the Web and for the portal's diverse audience. The launch content had to be developed in an aggressive timescale of four months and the activity required the team to create 250 documents from scratch. Each document was built around the needs of practitioners to make sure that the portal would provide short, readable and punchy documents, not long and inaccessible reports.
Collaboration across disciplines is very much a mark of successful neighbourhood renewal. The project demonstrated the benefits of this approach in microcosm by bringing together NRU and PA colleagues within the joint team. Bill Feinstein, Senior Knowledge Adviser at the NRU, says: ‘I found it really valuable to be effectively part of the project team. I felt that we avoided many potential delays or wrong turnings by working in this way. The assignment has felt like a joint endeavour; the spirit has been collaborative.'
A recognizable, memorable brand helps to ensure that the site's material appeals to those leading neighbourhood renewal programmes. The team developed a distinctive brand identity in collaboration with users and with the Office of the Deputy Prime Minister's (ODPM) communications team.
A trip to renewal.net
Renewal.net is now the authoritative source of evaluated evidence of what has, and has not, worked in neighbourhood renewal. At the time of writing, there are 1,878 documents available on renewal.net, including 277 evidence-based case studies. The collection continues to grow as more evidence is collected and linked.
The site's layout is logical, uncluttered and devoid of distractions. A master section introduces neighbourhood renewal, including overviews, case studies, explanation of various policies, summaries of research findings, FAQs (frequently asked questions) and links to other relevant sites.
The team created a simple navigation structure to help users find information quickly within a complex web of government policy. Renewal.net addresses over 130 renewal issues, such as attracting investment in land and property, reducing domestic violence and reducing truancy, yet users can go straight to the area they need. The navigation structure helps to make sense of government policy, demonstrating how the various initiatives ‘join up' to aid those involved in improving quality of life for communities.
New users can move to a ‘how to' overview introducing people to bringing about change in their neighbourhoods. The material is full of proven ideas for getting things done including: how to build a partnership; how to define the problem; how to select a project; how to implement a project; how to monitor and evaluate; and how to influence mainstream programmes.
The portal includes a comprehensive ‘jargon buster' covering the plethora of specialist terms used throughout the renewal sector, from ‘active community unit' to ‘working together learning together', complete with definitions and links to relevant sites. This resource in itself provides a handy primer for those getting to grips with the help available in neighbourhood renewal.
Users of the site can register for regular e-mail updates, detailing the latest additions to the site that are relevant to their particular interests. More than 500 users have registered to take advantage of this facility. The feedback from the user survey has demonstrated that this is an exceptionally valuable area of the site as it allows users to gain an understanding of new areas and how these reflect government policy priorities. Users can also access lists of upcoming events, searchable by region.
Elsewhere on the site, regional pages are being added gradually, so that users can quickly access documents relevant to their area. Entering ‘parks' as a search term from the London regional page in February 2004 fetched 363 documents in total, with 15 highlighted as being specifically relevant to the London area. The full document set included a case study on Emerson Park Development Corporation in East St Louis, Illinois, an overview of vehicle crime, research into the design of English seaside towns and an event about engaging refugees and asylum seekers in service development.
While the portal's main function is the provision of authoritative data, the site also acts as a community resource via its discussion forums. Anyone can browse the forums, and those who register can also contribute. Forums are organized around the neighbourhood renewal themes of housing and environment, worklessness, education, crime, health and local economies, as well as by region, practitioner group and ‘how to' topics. According to one user, renewal.net ‘is an outstanding Web site, and an invaluable source of information for me.'
The portal's effects
Renewal.net was launched in October 2002 by Barbara Roche MP, the government minister responsible for neighbourhood renewal. In the three months following its launch, renewal.net was accessed around 530 times every working day, giving an overall total of around 35,000 sessions. During an average session users looked at five different pages, giving a total of around 176,000 page views. The visitor statistics are nearly 50 per cent more than the targets given in the project's business plan targets.
More importantly, a recent survey has shown that the site has dramatically changed user behaviour, with 86 per cent of users indicating that the site enables them to take a more evidence-based approach to the design and delivery of renewal projects. Prior to the development of the portal, only 11 per cent based their project decisions on evaluated evidence. The growing use of such evidence will increase the likelihood that renewal programmes will achieve their intended results, whether they are tackling violent crime or graffiti.
Renewal.net is also helping policy-makers to adopt an evidencebased approach. For example, criminal justice bodies have begun to take a more evidence-based approach to drug abuse, which has led them to create links with the education and health sectors. A formal ODPM evaluation of the project and portal is currently under way.
The concept of the Internet portal has developed from that of a consumer destination to a corporate management tool. The NRU has evolved the concept further, making the portal a tool of the community. The earliest portals attempted to concentrate useful services in strongly branded home pages, with the intention of driving large volumes of Internet traffic through those pages. Such commercial portals hoped to derive advertising revenues from the many millions of ‘eyeballs' they believed would come to them first for all those people's Web needs. Most such portals failed because their brands and services could not deliver as much value as the comprehensive, optimized search engines such as Google or the authoritative news providers such as the BBC. In the enterprise sector, portals have been developed to guide users through the increasingly complex mass of documentation created for employee use. In many large organizations, corporate portals have replaced paper publication processes in areas such as employment policy or health and safety regulations.
The NRU has been able to exploit the development of portal sites and the underlying technologies that underpin them in order to serve a previously ill-supported, widespread population. Renewal.net does more than collect together trustworthy, usable information and make it accessible. The portal is also a major tool in the forging of a broadly based renewal community, providing it with structure, support and cohesion.
The project's sponsor, Barbara Roche MP, is clear on the project's fundamental aim of connecting experience with action: ‘For the first time ever, evidence on how to improve health, housing and education; how to create jobs; and how to reduce crime, has been brought together in a clear and accessible form, targeted at those doing neighbourhood renewal on the ground.'
The NRU has commissioned evaluations of the site's impact on specific neighbourhood renewal programmes. In the meantime, anecdotal evidence suggests that this project is helping the NRU in its goal of eliminating the gap between Britain's most deprived areas and the rest of the country during the coming decade.
PA managed to combine relevant business sector experience in urban renewal with the technical skills necessary to get a complex portal up and running in a short time. The PA team's insistence on document usability and attention to the crafting of each piece of content by the right experts shows an insight into the enduring value of information: it is only when people can find, understand and be inspired by relevant information that they can take action and make the changes they seek, whether those changes are in the sphere of business, the community, or personal life. The lessons of what makes content work on the Web, and how Web sites can best be structured for a wide variety of uses, are derived from wide experience of creating solutions in a number of different settings, both commercial and public sector.
This broad knowledge of how content is sourced, refined and consumed for maximum value is a valuable commodity for those consultants who have acquired it in their work with clients. Just as those citizens and officials active in the neighbourhood renewal movement share their knowledge of what does and does not work, so the best consultancies act as a means of disseminating and applying best practice throughout the organizations with which they engage. In this case, PA combined a commitment to plain English communication with systems development excellence to prove that the most far-flung pieces of knowledge can be brought together, made to make sense, and empowered to make a difference to ordinary lives.
Implementing workflow and document management technologies has given one of the world's largest insurers a blueprint for creating change across the business.
Aon is among the world's largest insurance broking and risk management companies and its UK operation, with two London offices, is a specialist in highly complex insurance risks. The London Market has a long tradition of face-to-face interaction, which is a complex, time-consuming and expensive way of doing business. Automation is now catching up with the London Market, and players are beginning to modernize their business processes as the market abandons its reliance on paper mechanisms. The overall evolution of the market is governed by its progress in enabling electronic transactions. However, individual players can make important steps towards a paper-free environment by reassessing their internal processes. By addressing their internal needs they can make savings and create flexibility, while making themselves better prepared for the dawning era of inter-company electronic trading.
In 1998, Aon Limited decided that its long-term future lay in greater electronic trading with its business partners. Electronic trading would demand more efficient ways of exchanging information, both within the market and within the company itself. The company realized that the change to electronic trading would put great strain on the company's systems, business processes and staff.
The move to electronic trading would clearly need a new system to improve the company's handling of both insurance content and transactions. Aon began a four-year partnership with Impact Plus to deliver an environment that would yield savings in its business processes and make the company ready for the online business regime emerging in its industry. Impact Plus's Graham Whitehead says:
The way they were looking at it was that they wanted stepping stones to electronic trading. They wanted tools that would allow them to understand where in the renewal process a particular risk was, what correspondence they had received from a client in relation to a particular risk or set of risks, and to get a clear picture of their risk processing responsibilities.
The team began by building a seven-year business case setting out the value of introducing document management systems, workflow technology and new ways of working to go along with the new support systems. Over a period of three months, Aon and Impact Plus staff worked together on the business case to gain a thorough understanding of the way the company worked. To ensure ownership within the business, the business case was discussed at every level within Aon Limited, and was then presented to executives from Aon's headquarters in Chicago.
The key to the business case's approval in 1999 was its focus on automation as a lever for greater control of the business. New systems would help achieve a far greater level of process consistency across the business, while improving customer service by providing fast access to client information. Savings would be made by better managing the escalating amounts of paper documents and e-mails being sent, received and duplicated. At the same time, improvements to the company's systems would allow it to embrace industry initiatives on Regulatory Compliance and the exchange of important information among industry participants. It all adds up to the largest project Aon Limited had ever undertaken at that time, and which is on target to save the company tens of millions of pounds while readying it for improved competitiveness in the evolving insurance industry. Achieving the project's goals will give Aon a level of efficiency and organizational flexibility not found in any of its competitors.
Commonality and differences
Though document management and workflow systems will improve the productivity and flexibility of any information-intensive organization that adopts the technologies, these are not simple generic solutions that can be successfully applied to businesses in a random fashion. While the broad benefits of generic applications such as word processors or Web browsers make them relatively straightforward purchases for the organization, workflow in particular strikes right to the heart of what the business does, and how it does what it does.
Workflow systems formalize and support the processing of packages of information within multi-disciplinary environments. Workflow is the white-collar world's equivalent to process automation via conveyor belts and specialist workstations. In a typical workflow system, sets of information such as funding applications are captured electronically on their entry into the organization. Documents that arrive in paper form are scanned and stored as digital images, and may also be converted into computer-readable text using optical character recognition (OCR) techniques. Incoming documents are categorized and tagged, much as traditional paper-based files are created.
But since a workflow system lives in a networked environment, electronic files can be sent immediately for the attention of the staff members that need to work on them next.
Workflow systems allow document collections to be routed in sequence or in parallel, depending on the underlying business process designed into the system. So, for example, our funding application might be routed to three individuals in sequence so that each can check different aspects of the application for completion of sections, adherence to the rules of the scheme and supporting documentation. If the application passes these three stages, it may then be passed simultaneously to the members of a decision-making board for their approval. The workflow system marks the changing status of each case it deals with as work is done on the case. The system therefore acts as a combined library, document control mechanism and progress-chaser.
While some technologists have attempted to develop generic workflow processes for many business domains, including insurance, these have rarely met with success. One reason is that even the best generic system requires tailoring to its individual target business environments. Tailoring the generic system entails researching the business processes of the target organization, just as creating a system from scratch would do. The task of adapting a generic framework to particular environments often has no cost or time advantage compared with starting with a blank canvas.
Another reason why generic workflow processes meet scepticism lies in every organization's firm belief in its own uniqueness. Companies understandably believe that their success to date indicates their competence: they have grown their own businesses, and see their processes as an organic part of the success they have achieved. Furthermore, the mechanics of competition suggest that success can be related, if perhaps only obscurely, to the manifest differences found in each player's approach to business. It is seldom possible for management to point to a particular aspect of the organization's operations and associate it with business dominance, and when such an identification is made the focus is often on a novel, standalone system imported from outside the organization or developed in-house as the dream of a single motivated individual. In traditional white-collar environments such as the insurance industry, where inputs, outputs and value-adding processes are all composed of information, the connection between process design and business performance is especially hard to trace. The result is a paradox for managers tasked with creating change: people tend to believe in workflow as a concept, but believe equally strongly that it cannot be applied in their own environments.
Phil Tyson, Aon's Programme Director for the project, recognizes that these beliefs are intimately related to the company's business domain: ‘The highly specialized area of insurance in which we operate has a 300-year heritage of doing business face to face and a track record for resisting change; we knew that the cultural hurdles would be significant.'
Aon is a good example of the problem of turning from the generic to the particular, since the company exhibited all the variety of its industry within its own walls. Each business unit performed similar functions, but each was convinced it worked differently, and that the differences among divisions were essential. The team found that the practical consequence of Aon's evolved habits were complex processes involving much shuffling and shuttling of paperwork between teams, and a heavy photocopying bill. In order to add a new insurance policy to the various systems in the relevant business units, a key contractual document had to be photocopied at least six times. The team also noted that years of inertia has helped make the current inefficiencies appear normal, and created a reluctant attitude to change. If technology was to be a business enabler at Aon, the business environment would also need to be addressed to ensure that the technology solution ‘took'.
Twin tracks to change
The Aon/Impact Plus team ran a vendor selection process to find and engage the right technology partners. Tenders were invited from selected consortia and the responses scored against pre-agreed criteria. Reference visits were then made to large reference installations around the world in order to see the various types of solution in action. A winning consortium was chosen based on its knowledge of the London Market and the proven quality of its product set. This consortium was made up of FileNET, a leading provider of content management and workflow tools, and systems integrator August Group. The team then turned to the task of creating the changes in culture and working practice it had identified as crucial to the project's success.
The team invented a two-pronged approach to the project. One stream of activity would concentrate on IT delivery, or ‘getting the technology ready for the business', as the team defined it. A second, parallel stream would work on internal change, ‘getting the business ready for the technology'. This approach was based on Impact Plus's Programme and Change Management (PCM) technique.
The members of the business readiness stream worked closely with Aon staff to prepare the 11 business units for the coming electronic revolution. The workstream created ‘change teams' in each business unit, meeting on a weekly basis to plan the phased introduction of new working practices. The change teams were supported by cross-departmental analysts, who were able to identify and communicate similarities and differences among the different Aon teams. There were over 2,500 people in the business units, each specializing in different areas such as aviation, shipping and financial services. All these staff believed that their working practices were necessarily unique. The team was able to show them that many of their processes were very similar and that, with some redesign, they could benefit from following common processes.
Getting people to recognize similarities in processes can be difficult to achieve; but once those involved have begun to represent common processes, they can rapidly move to the opposite extreme and begin to search for ‘the one, complete system' that will work for everyone. The team knew that this is a misleading path. According to Whitehead: ‘The typical route is to define processes very tightly, but that way it gets too complicated. Our view was: we've got the experts, but the administration tasks are holding them back. Let's let them take their decisions, rather than try to put all their expertise in the system.'
In other words, the team's guiding principle was to target the mechanisms of the work, not the expertise behind the work. The aim was to remove paper, not brains. The project's clear respect for the value added by the staff was a major aspect in its success. Requirements for the system were gathered diligently from everyone involved. Each view of the system's requirements was captured, documented, shared and challenged.
The business change workstream
The team set out five goals for the business change workstream:
Create a carefully phased, low-risk introduction of the system to the business.
Establish clear ownership of the new way of working by each of the business units.
Create a well-planned change process.
Ensure that common business processes are followed by all the business units.
Establish a clear focus on achieving the benefits within each business unit.
In order to phase in the new system, the project team worked with each change team to understand the business unit's trading cycle, the pattern of peaks and troughs in its workload and the competitive pressures it was experiencing and predicted for the future. Project activities could then be planned to fit into the business cycle, minimizing the impact on the continuation of normal business. The team also set up a dedicated ‘familiarization' area so that change teams could get hands-on experience of the system as it developed. The familiarization facility allowed staff to explore and discuss different ways of using the system, and to choose those that would generate the highest level of benefit. Tyson says: ‘Feedback from the familiarization sessions proved that this approach reduced the risk associated with going live and helped greatly in removing the uncertainty and even fear associated with the new system.'
Early access to systems helps to show staff that what is being built is intended to be relevant to their duties and preferences. This kind of exposure also helps trigger ideas about better ways of working, redundant procedures or areas of ambiguity. From the point of view of a system's functionality, many people find it hard to specify what they want a system to do in the abstract, but are very capable of critiquing a working prototype and using their reaction as a springboard to defining their requirements. This experience also demonstrates to those staff involved that they are genuinely in control of the development, and that the business benefits of the system really do form its top priority.
Ownership of the project by the business units was designed into the project's delivery by the creation of the change teams. Each change team grew steadily in capability and confidence to a point where it became wholly responsible and accountable for delivering the necessary change within its own area of the business. The change team, rather than the project team, became the first port of call for queries from staff within a business unit. Change teams also took ownership of the communications exercises linked to the roll out of the system.
The business units had only minimal experience of working with formal project planning tools and were initially unconvinced that change could be delivered successfully. Staff views were informed by the prevailing belief that technology had failed to deliver against its promises throughout the industry. In order to reassure each business unit about the organization's determination to achieve its goals, pragmatic and detailed project plans were created, based on a common template created by the project team. Each plan showed the key deliverables and milestones for the unit in a single page. This concise presentation focused the business change activity around specific goals and outcomes, and demonstrated that the project was under control.
The team was also able to identify parts of the business processes that were common to all 11 business units and where common workflows could be followed. These processes were then automated using the FileNET workflow tool, and became the core of the new system.
Highly visible planning, ownership and process modelling activities demonstrated the team's commitment to readying the business for a step-change in performance. All of these activities had also to be coordinated with a clear focus on achieving the business benefits determined by the project's business case. Widespread and regular discussions about benefit realization therefore started early on in the project. Each business unit was made aware of the timescales and level of benefit expected for its area.
Organizations always experience a reduction in productivity as users get to grips with a new system, but the team did not want the changeover to undermine the realization of the project's business goals. The project team therefore guided each change team in quantifying the scale and duration of the expected dip in productivity, and helped each business unit to plan activities to minimize the impact.
The change teams ensured a gradual and orderly transfer of ownership from the project team to the business. This approach also allowed Aon to keep the project team to a small size. The new flexibility and capability for change proved critical, in the aftermath of 11 September 2001. The attacks on the World Trade Center led to some of the toughest trading conditions ever experienced in the London Market and the change teams played a vital role in ensuring that, despite the extremely complex business environment, the implementation continued to progress through this difficult period.
The systems stream
The design and build of the new system took 18 months, with new releases of software and working practices issued at regular intervals. By delivering the system and its working context incrementally and in lockstep, the team was able to transition the business in an open and coherent manner, with improvements to functionality arriving in discrete and digestible portions. This strategy echoes the ‘stepping stones' philosophy underlying the project.
Among the key criteria for the project's success was the reliability of the systems it delivered. Aon staff pointed out that when their existing systems were unavailable, they could revert to a paper back-up. But under the new plan, which involved a wholesale shift to electronic files, if the system went down they ‘might as well go home'.
The team therefore designed a highly resilient system with a standby server and a disaster recovery system at an alternative location, to ensure that users would have near-uninterrupted access. The standby system can be brought online within 20 minutes if a component in the main production system fails. The disaster-recovery server can be made available within a few hours, should a disaster affect the first two systems. The system architecture has been designed to allow for automatic and instantaneous switching between any of the three systems in the future, should the need arise. This is an example of planned replication, undertaken to ensure resilience. Aon used to replicate procedures because it had no better way of performing its business processes. Now it has standardized its business processes, but replicated their support mechanisms to ensure continuity of delivery.
To date, the computer system and associated new ways of working have been rolled out to more than 2,500 users. The business now has automated processes that are fully compliant with industry regulations and business standards, and open to audit. More efficient ways of handling correspondence and policy-related documentation have been installed, and all users have access to common, well-understood business processes. The company has achieved endto-end automation of new business, administration, claims and accounts processes. It now has a single repository into which all documentation - including e-mails - can be filed and to which everyone has access. Links to Aon's policy documentation and correspondence production system, central processing system and e-mail system are in place. The new systems environment also includes a single authoritative reference list of client organizations, replacing a plethora of previous databases. Populating the new client database involved purging the company's aggregated records of some 30,000 redundant entries.
Change projects measured in years inevitably have to roll with the punches delivered by unpredicted changes in their external environments. This project had to endure a great challenge to its programme when the terrorist attacks of September 11 led to the rapid re-engineering of the company's very industry. The challenges of electronic trading in its market and inertia in its organization were dwarfed by a sudden change in the insurance industry's posture in the face of increased global terrorism. However, the commercial effect of the disaster was an upturn in premium trends. The industry's cost-cutting attitude switched to one of growth. A key implication for Aon of the industry's change in direction was a shift in emphasis from reducing staff to retaining them, and their valuable expertise. This project is therefore delivering on its business case in a different format from that originally envisaged. The business benefits are being measured in terms of improved service, quality and capacity rather than cost reduction. Another way of looking at this change of emphasis is to say that the project has allowed Aon to direct its investment towards growth at the required time.
Impact Plus introduced a key principle into its work with Aon: the twin-track approach to change. The claim that ‘technology must serve the business' is often made by consultants but undermined by single-minded attention to the installation of systems, and an unthinking disregard for the working environment in which those systems will function. Impact Plus was careful to create equally weighted streams of activity aimed at creating the best possible climate for the introduction of the new systems. The firm was also determined to give equal billing to the new ways of working that would surround the new support functions, so that the human dimension of the change never fell from view. By pursuing their twin-track approach, Impact Plus proved that, in business, parallel lines can indeed converge on a successful implementation.
‘Implementation' is another keyword to associate with this project and Impact Plus's general approach to its work. The firm has a strong belief in implementation, not just ideas. For this team, workflow is a lens through which processes can be examined and corrected. It is detailed, nuts-and-bolts work that does not respond to grand gestures or sweeping visions. This attitude gives the consultancy a highly practical take on its own capabilities. For example, looking back on the project, Impact Plus's Graham Whitehead finds one aspect he might do differently, if he could turn the clock back: ‘We could have done more role play - more scenarios. We ran scenarios in the familiarization environment, but that was well into development. For example, 'client' can mean different things to different people, and we found out through scenarios that we needed a richer definition for it.'
The lasting effect within Aon is the recognition that change is possible, and liberating. This project has given the company a blueprint for making change, based on a dual-stream approach to technology development and business readiness. It is an approach that is now being used on other projects within Aon as the company continues to respond to its evolving business environment.