Technology dooms dotcoms

A bricks-and-mortar company might have survived the problems that sank Boo and Toysmart, but for e-tailers technology's...

A bricks-and-mortar company might have survived the problems that sank Boo and Toysmart, but for e-tailers technology's limitations can undermine the business vision

Duncan Chapple


The recent failures of online businesses and have serious ramifications for e-tailers, whether they are start-ups or established companies going online.

Both Boo and Disney's Toysmart were doomed by their inability to get the technology to deliver the customer experience they had planned. They found that there is a lot more to running a dotcom company than the technology. But an e-business that can't get the technology right is an ex-business.

A bricks-and-mortar company with similar problems could have recovered. Customers cannot always see when an offline business has inefficient technology; online, it screams out. Businesses based on one "killer" feature - such as Toysmart's personalisation or Boo's visualisation - get killed double quick if they can't deliver.

Boo wanted shoppers to be able to manipulate 3D images of merchandise. The problem was they couldn't deliver their business vision. Between Boo's overloaded Web servers and most people's keyhole-sized Internet connections, those images simply could not get through at a usable speed.

It was Disney's backing that made Toysmart one of the hopefuls of US e-commerce. One of the first online stores, it aimed firmly at the high-value market for quality educational toys. After benchmarking its competitors and listening to US market analysts the company decided to build a highly personalised Web site.

In discussions with staff at Toysmart in the period before its closure in late May, I was reminded of some of the make-or-break lessons that e-tailers learn the hard way. My contacts at Toysmart found that Microsoft Site Server was an easily administered and inexpensive entrance into online commerce with some useful personalisation features. However, like many businesses, it found that Site Server's scalability was certainly not linear. As the business grew, adding extra servers increased power with incrementally less impact.

While constantly tweaking bottlenecks through the winter shopping season, Toysmart went from eight servers to 35. Time spent firefighting meant there was no time to beef up the site's personalisation features. Yet personalisation was the make or break feature needed to make the site sticky enough to win return custom.

Many e-businesses rush online or hurry to add new features to their site. But without proper load testing, a strong and secure architecture and flexible relationship e-commerce software, e-businesses are stuck with wheezy, boring out-of-the-box Web sites.

E-business software consultant Duncan Chapple is co-author of Ovum's report eCRM: Personalisation Technologies for the Web.

E-business: lessons from the front line

  • Get the business vision right. If your killer business model is not supported by existing software, think seriously about building software to enable the vision

  • Implement features incrementally, only after you are sure that the scalability, security and architecture exists to support them

  • Sticky sites need complex personalisation and customer relationship management

  • The biggest technical challenge is complexity. Business-to-consumer commerce involves rapidly changing content and business rules. Simple off-the-shelf Web shops can freeze your business, so look for process workflow, e-customer relationship management and Web content management software that gives good support to changing business needs

  • Don't take the supplier's word for it - or the reference site that got the software for free. Try software before you buy it

  • Don't build everything from scratch, use existing software and pre-built components where you can. Source: Ovum

  • Read more on Business applications