So, recent research from TNS/Zeus Technology shows that 82% of UK online shoppers said that if a website performed badly it would dissuade them from buying goods from that organisation on the web or in store.

The survey was based on a sample of over 2,000 shoppers and shows that each of them will abandon around 5.5 online purchases every year due to poor website response times.

Multiplying this by the average value of the shopping basket and the predicted number of online shoppers we can calculate lost sales of around £5.65bn a year. Not to mention damage to the retailer's brand - both online and offline.

If this is the case, shouldn't someone be tackling the website performance problem?

Some of the performance issues are due to playing bandwidth-intensive videos for advertising purposes, and surges in the number of simultaneous online shoppers (eg during sale time). Other performance issues are due to technical bottlenecks such as:

  • Application design
  • Server configuration
  • System architecture
  • Poor system integration

The sad thing is that poor performance is entirely preventable. Performance testing is a well-established process and powerful new test tools are available and affordable. NeoLoad, for example, can simulate thousands of users using a website in a variety of ways. Its reports will show exactly where bottlenecks are, and show what the limits of the website are.

A cynical view would conclude that web performance issues exist because solving them is not considered important enough.

But given the predicted loss of sales, performance testing should be high on the agenda for not only the IT department, but also sales and marketing.

For information on how to use performance testing to improve your customers' online experience