The failure of William Hill's Nextgen online betting platform, which was developed in-house, raises questions about whether IT directors should embark on business-critical software development projects in-house or use third parties.
By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
In a trading statement issued earlier in the month, the company said it had decided to abandon its Nextgen project.
William Hill said that following an independent review of the project in November 2007, the board, "decided to terminate the NextGen technology programme and to implement an externally developed third-party technology solution". The failed project will cost William Hill £26m and the replacement system is expected to be implemented by the end of the year.
The project's objectives were to move William Hill from legacy IT on to a modern IT infrastructure using a Java-application architecture based on Java, BEA Weblogic/Portal and Oracle. The migration was designed to make William Hill more agile and enable it to offer new services to its customers, which would differentiate William Hill from its competitors.
With so much at stake, why did it fail?
Any IT director running or considering running a major software development project should be aware of the limitations of internal development work, said Gary Barnett, partner and chief technology officer at analyst Bathwick Group, "There is a massive scale issue between creating a relatively small in-house application that is well-defined versus and creating an application that runs your whole business."
The difference in scale makes enterprise-wide projects exponentially more difficult to get right, according to Barnett.
With this in mind, Barnett urged IT directors to consider whether they would be prepared to invest in the level of source-code management, project management methodology and skills of a commercial enterprise software company. Without tight control, Barnett said internal projects risked growing in complexity and ultimately failing.
Neil Ward-Dutton, research director at IT advisory firm MWD, said consumer-facing websites tended to be at the forefront of using new technologies. "It can be quite easy to jump on the 'me too' bandwaggon without looking at why customers need the new functionality."
Online betting sites have a unique challenge. "They have to handle a large volume of traffic and are handling people's money, so site processing needs to be in real-time to avoid changing the betting odds," Ward-Dutton said. "The transaction rates they have to sustain is like a small financial market handling thousands of transactions per second." At some of the sites Ward-Dutton has looked, the IT team has needed to rewrite application servers to make them more efficient and deploy in-memory databases to support the peaks in betting traffic.
But all businesses should focus on tight software development. It is not uncommon for IT directors to work with an external consultant to fire fight when a project starts going wrong. The experts agree that prevention is a better approach, and IT directors should consider using such consultants to help them establish basic software development project management.