One year on: e-business pioneers one year on

Don't rest on your laurels after a successful launch. It's review time! Helen Beckett tracked down four e-business pioneers to...

Don't rest on your laurels after a successful launch. It's review time! Helen Beckett tracked down four e-business pioneers to discuss the lessons learned one year on from launch

The Bank

smile.co.uk

Interview with Keith Girling, technology director for the Co-op Bank

Smile is part of the 127-year-old Co-operative Bank plc that operates on ethical principles.

The business case

The Co-op Bank has a multiple channel distribution strategy including call centres, PC, TV (over the Sky network) and mobile phone. We introduced an Internet banking channel in March 1998. Set-up costs aside, it is estimated that a pure online bank costs one tenth of the sum to run a branch operation, and one quarter the costs of a call centre and we wanted to pass some of these savings onto the Internet customer.

Project scope

It had to be a complete system connecting the customer PC through the Internet cloud through the middleware layer to the host banking systems. Some of these back-end systems had to be created from scratch.

Implementation tips

We followed the bank's strategy of using no internal IT development whatsoever. The entire project was put together using 40 external suppliers from the UK, Europe and the US. This meant we could be flexible while retaining "currency" - I'm only interested in next week's skills. However we, as usual, retained integration and project management in-house.

Lessons Learned

Our ambition was to be first to market and so we were prepared to sacrifice some things to achieve that goal. If there was one thing we underestimated it was the extent of the overhead of having to go back to make things fit in more elegantly for the future.We learned how vital it is to keep thinking about an overall strategic [IT] architecture.

Business assumptions revised

We're so close on all our key metrics that it's a bit embarrassing. We were able to model on customer take-up and performance based on our previous experience of call centres roll-out.

The Bookie

willhill.com
williamhill.co.uk

Interview with Ralph Topping, online betting director

William Hill bookmakers can be found on most High Streets - there are 1,550 outlets in the UK.

The business case

The company has always been aware of the advantages of exploiting technology to make betting more convenient and accessible. In the 1990s, William Hill introduced telephone betting for debit card punters. This rapidly accounted for 20% of revenues and the company launched a brochureware site three years ago to try and funnel more business through the telephone/direct debit operation. The real incentive to build an e-business happened in early 1998 when the MD took me aside and said, "There's a World Cup in the summer - I want a transactional Web site ready for the customers to place a punt in the summer." There's nothing so motivational as getting a good kicking from your MD.

Lessons learned

The IT team rushed out and bought kit, the idea being that we would replicate the telephone-based booking system that was so successful: that was our first mistake. The kit was Stratus fault-tolerant servers, but the team realised that speedy response was much more important for Internet business and switched to HP servers. At the front-end, our original brochureware site had been very graphics-rich. We thought that it had to be pretty to encourage people to place a bet - but again speed proved much more important. We used two in-house designers, who could retune much more rapidly than via an external design agency: that decision saved us £200,000.

Implementation tips

If you have in-house technology that works for your business, then use it. William Hill over the years has developed a post-relational database that is a flexible way of storing and searching for data. It was ideal for e-business. A lot of companies don't have the expertise, drive and ambition to use in-house systems - they bring in a third party and put all their eggs in one basket. Make sure that the programmers and designers and strategists are instantly accessible to each other. You need to eyeball your programmers so they're not relying on interpretation over a telephone line

Business assumption revised

None really. The Internet has opened up a hot market in south east Asia which now accounts for 90% of William Hill's online business. Our aim was always to make a lot of money - we're not a dotcom that plans on breaking even down the line.

The Motor Insurer

Ironsure.com

Interview with Davis Shelley, general manager Iron Trades.

Part of Iron Trades, one of the top 20 general insurers in the UK, which for the past 100 years has provided a portfolio of services

The business case

The arrival of Direct Line in the mid 1980s with its direct telesales model transformed the insurance industry - and we missed out on that opportunity. In 1998 we decided that the Internet was going to be big, and that we would start early and differentiate ourselves from the crowd.

Project scope

We wanted to be a pure Internet business and that meant not just buying a policy online but being able to update it and perform all the administration online too. A lot of names were taking existing functionality and grafting it onto the Internet - we believed that some of the online forms being presented would give the Internet a bad name.

Lessons learned

We underestimated the effort and [continuing] expense of building an online brand. Being first and the best in the market is no guarantee of success. How much to spend and where to spend it has been a continual learning curve - we have spent several times more than we originally anticipated on branding.

Implementation tips

We started with nothing other than a legacy mainframe system, and that proved a blessing. To begin with, we assumed we would have to build the entire system - front and back-end - from scratch, especially as the mainframe has to be switched off once a month for maintenance. However, the discovery of middleware (MDIS Enterprise Access) opened our eyes and its ability to make the back-end engine serve new front-end functions.

Business assumptions revised

Basically we got it right - the Internet was the right medium to back. Although it's been slower to take off in the insurance market than I'd expected, the adoption of the Internet for insurance products is at the beginning of the curve. The possibility of acting as a wholesaler and providing the online brand to an existing name is still a potential business model for us.

The record distributor

hmv.com

Interview with Stuart Row, general manager, HMV e-commerce

The business case

We knew we had to start selling online when we were getting around 500 e-mails a day from our customers asking us to make stuff available over the Internet.

Project scope

The company decided to launch its e-business strategy in two phases. A "quick fix solution" was launched in 1997 - based on Microsoft SiteServer 3 and NT - to tide the company over for a year while it developed a robust, scalable platform based on IBM hardware and software.

Lessons learned

The one year "pilot" was very useful in that we realised that back-end fulfilment was crucial and also that it was essential to offer a wide product range. Another issue we had to resolve was how to deal with the massive amount of e-mail. At the time there were only a couple of products on the market to do the filtering and we brought in Egain to do this.

Implementation tips

Hire people who have a sense of humour. There can be really exasperating times when things go wrong and you need tenacity to get it fixed. It's an unpredictable environment and the growth of e-commerce means that projects tend to be under-resourced. Also don't get too obsessed with a launch date. Even though you always launch with problems, you can launch too early.

Business assumptions revised

We knew there would be a lot of back-catalogue demand but didn't realise just how deep or specialist this would go. HMV's biggest store on Oxford Street has 140,000 products - our Web site has a quarter of a million.

Read more on Business applications

SearchCIO
SearchSecurity
SearchNetworking
SearchDataCenter
SearchDataManagement
Close