You are here  Software

One big happy family

David Bicknell
Tuesday 24 July 2001 04:38
Integrating diverse Web applications, especially in this age of mergers and partnerships, is no easy task. David Bicknell looks at how Royal Sun Alliance has tackled its integration dilemma

Eighteen months ago, as traditional companies began getting to grips with the Web, their various departments were all encouraged to build Web applications; to play around with and learn to understand the technology. So lines of businesses all developed Web applications. They all planned and budgeted separately, exposing their departmental systems using different development tools and isolated databases. This approach benefited neither the customer, nor corporate management.

For the customer, it meant often having to register details separately on different parts of a portal, or having to log-in to different parts separately, with different user names and passwords. It often meant having to repeat personal details, such as an address, when trying to get buying details or quotes.

For corporate management, there was no management information - no ability to track user visits, activity and paths across the company portal as a whole. Nor was there the ability to track visitors on any partners' sites or cross-sell through knowledge of customers' navigation paths through the portal. The poor linkage to call centre customer relationship management systems affected customer service, and overheads increased as a result of having to manage multiple Web development projects with different technologies and skillsets.

This is a common problem across all industries that have had to embrace the Web. The creation of what are known as separate 'silos' has meant that the majority of Web portals offer little transparency, and little meaningful management information.

One company that faced precisely these problems recently was UK insurer Royal & Sun Alliance (RSA). Having gone through a 1996 merger between Royal Insurance and Sun Alliance, it needed to integrate the two companies' product lines. Although it had a direct presence, offering separate insurance products for the home, motor, travel, life, health and investments sectors, the insurer wanted to create a lifestyle package, thus integrating all these separate insurances into an easy-to-use solution for the customer, which in turn would yield more business.

What was equally important to RSA was creating an online brand that would be synonymous with excellent service, giving customers complete choice of communication channels and always going the extra mile to solve customer problems. But it realised that opening up all its existing systems to customers also meant opening up all its foibles to customers as well.

Currently, RSA has about two million customers using only one RSA product each. The company wanted to increase its customer numbers by 400,000 a year and by 2004 wanted to take that customer holding up from one to three products. It needed to ensure that the system facilitated compliance with financial services rules, that it could get new products to market quickly and that it could add sticky content equally quickly.

Some of RSA's insurance offerings are co-branded with partners such as What Car magazine. But RSA was frustrated that despite spending money to buy content through online partnerships it was unable to measure the spend's effectiveness. It wanted the new brand to solve that issue, too.

An additional problem for RSA was that it wanted to develop its new brand ultra-quickly, needing to get the brand out before the summer slowdown. After appointing a new marketing director, Mike Tildesley, to develop the brand, dubbed 'More Than' - an exercise which took from September to January - the company had just five months to deliver the project. RSA insiders suggested that they did not have enough time to implement the project.

There were organisational and technical problems associated with delivering the new brand. The organisational challenges included becoming more risk-tolerant and making failure acceptable; consolidating the work of 11 internal and 15 external IT departments; doing parallel development; having clear accountability and regular communications; and making sure that project stakeholders were able to make decisions on-the-fly.

The technical problems included different silos of technology for different insurance products; different silos of technology for different customer touchpoint technologies, such as PC, WAP phone and digital TV; a different look and feel embedded in each silo; no consolidated content management and no consolidated business intelligence.

According to Tildesley, the whole system was "organised entirely the wrong way for our customers and for us".

Having taken a decision to undergo the rebranding exercise, RSA was then faced with the problem of how to do it both quickly and cost-effectively. It had three choices. It could completely rewrite existing Web applications, but this approach would have been prohibitively expensive and time-consuming. Another problem was that the company didn't want to rewrite all that content since it was happy with 60% of what the existing Web applications already offered. So, rewriting was neither possible nor justifiable.

A second option was to colourwash the applications to give a common look and feel. This would have required piecemeal programming of each of the product-related silos incorporating Web applications because a look and feel was already embedded separately in each of them, with all using different suppliers, such as IBM, Microsoft and Macromedia.

Although such an approach would still probably have allowed RSA to hit its June target date, it would have increased dependency on skillsets in other departments. In addition, RSA wanted to gain portal-wide management information on customers and be able to offer up-to-date product information as well as the facility to offer access from non-Web devices such as IDTV and WAP-based mobile phones. A colourwash would still not have helped achieve those objectives.

The third option was that it could find another quicker, more practical approach.

RSA decided to adopt technology that, in effect, put a wrapper around the existing disparate Web-based applications to create a single view. InterX's Net2020 software, in essence a Web application integration package, was the only option, according to RSA, that would enable it to meet its rebrand timescales, balance benefits and costs, and ensure existing online applications did not need to be replaced.

In addition to adopting InterX, RSA also assembled a specialist team to drive the rebrand strategy, integrate the system and project manage the exercise. These included representatives from Ogilvy & Mather for brand strategy, PricewaterhouseCoopers for programme management, Ogilvy Interactive for Web creative, CMG Admiral for project and technical skills and London Bridge Software for call centre software.

In all, the exercise is likely to be budgeted at about £40m, split roughly 50:50 between rebranding and technical costs. Of the £20m or so being spent on IT, the InterX solution is estimated to cost about £1m.

Having got a number of partners on board to deliver the project, RSA also had to ensure it controlled progress. This was achieved by conducting daily meetings with all the parties, but at the same time giving executives the ability to make decisions without having all the information at their fingertips.

In future, Tildesley hopes the system will be truly multichannel, enabling users to choose the technology with which they contact RSA. Already, RSA is offering its clients a facility where they can receive updates on their claims by SMS text messaging to their mobile phones rather than via snail mail, although a written audit trail for those claims still exists and critical correspondence will still be paper-based.

The rebranding exercise, which has been accompanied by a high-profile advertising campaign featuring a dog called Lucky, is an example of what can be achieved in a short space of time. It is also an example of how traditional companies, having developed Web products hurriedly and haphazardly in different parts of their business, are going to have to find ways of ensuring that both their management teams and customers see an overall view of the business online, with a consistent look and feel.

What RSA has recently achieved, others will be eager to follow.


RSA's Rebranding Exercise
Before After
Fragmented brands One brand with coherent look and feel
Silo-based products Packaged lifestyle solutions
Slow, unique delivery Speed to market
Siloed channels Integrated customer experience
Lack of management information Full management information


Why Organisations Struggle with building silos
CAUSE:
Organisational departments and lines of businesses were encouraged to experiment with the Web, but their piecemeal building, or creating silos, left portals with a jumble of fragmented applications, built with different technologies, and an inconsistent look and feel.

EFFECT: These bits and pieces portals have left firms short of management information on the value and effectiveness of their online operation, and left customers having to register their details numerous times.

FUTURE: Web application integration (WAI), where products such as InterX integrate the silos by acting as a wrapper for them, are the future. The alternative to WAI is enterprise application integration (EAI), where companies re-architect their middleware layers, but this involves significant programming effort. One UK organisation with a number of websites wanted to reprogram them all with EAI tools. After spending about £4m, it has suspended development.


Web app integration
RSA has used InterX's Net2020 to strip out the presentation layer of the existing RSA and partner websites and create a single portal view. It uses page templates and components dubbed NetPages and NetElements to map to items on pages in existing Web applications, enabling representation for portal standardisation, rebranding or merging. Items in existing Web applications are tagged so that they can be read by Net2020.