Customer response is measured by metrics such as the conversion rate, total online sales, bounce and exit rates, etc. all of which can be found simply if you click on your analytics dashboard. But similarly, you can’t just rely on your analytics data for the entire picture. There are metrics that are based on the site’s performance, if measured correctly, can really help pinpoint the symptoms that are leading to low conversion rate etc..

According to Internet Retailer’s Andrea Mulligan there are 3 performance metrics that are directly related to the online experience and of course impact sales: Store Availability, Response Time, and Consistency. The Internet Retailer’s Top 500 were measured against these metrics:

1 – Store Availability

Nothing is more frustrating to an online visitor that is browsing through your site than to find that certain pages are unavailable. The average availability for the Top 500 was at about 99.71%; 104 sites had an availability rating of 100% while the lowest lingered at 99.5%.

Looking at this a bit closer if your site receives 1 million visitors per day, then 99.5% availability translates into 995,000 visitors are able to access the site while 5,000 visitors not being able to access your website on any particular day. I guess 5,000 might sound too little if you are dealing with million visitors a day.

But let’s consider how much lost revenue this translates into.

If your websites converts around 2%, then 5,000 visitors could have generated 100 additional orders (5,000 * 2%). Let’s say your average order size is $100,then your site lost:

100 orders * $100 per order = $10,000 per day or $3,650,000 orders per year!

2 – Response Time

The reason why people turn to the web for their shopping needs is because of the convenience, speed and ease of it all. So a slow-loading page and website will definitely impact the site visitor’s experience and may lead to less than expected sales. The average loading time for the homepage of the Top 500 was at 3.564 seconds and the top 207 e-tailers were faster than that. Only a handful of sites had a homepage load site of less than 1 second. The slower sites lingered at numbers between 12 – 18.434 seconds.

Response time has a direct impact on conversion rate because, generally, people who shop online are too busy to go to the brick and motor shops in the first place. Most online transactions happen at the work place or somewhere other than home, which means the potential customer is on the run and in a hurry.

There are many reasons why your website performance might be slow and I hope to covers some of these in an upcoming blog.

Again, this metric is one that is measurable and etailers alike should pay heed to it.

3 – Consistency

If you’re looking for stronger brand loyalty across the board you need to have a consistent site. That means whether your site visitor is in Seattle or Sydney and is checking the site at 1 am or 1 pm, the performance metrics and site in general must be consistent and stable which will translate into greater online sales and longer time on the site itself.

Consistency levels for a site are generally measured across various times and traffic levels. Among the top 500, 40% of them had mediocre consistency. Some companies that scored poorly when it comes to consistency yet experience already large volumes of sales can consider improving consistency scores to easily increase online revenue and conversion rates.

After reflecting upon this specific metric, I feel that it was non-conclusive and did not fall within the perimeters of performance metrics as we see it. Consistency whether you’re in Singapore or Orlando relies a number of factors many of which are out of the etailers control. So to place this as a performance metric is unfair to the e-tailer and inconclusive of its data.

Tying it all together, in order to measure site availability and response time, frequency of measurement must be a big consideration: A site on average should measure the store availability and response time on the hour. More successful sites go as far as measuring their metrics every 5 minutes making conclusive assessments and business decisions based on the results of this data.

Of course there are other performance metrics that we consider but are a bit too technical in nature to include on this blog such as user session size, memory utilization, database usage, and CPU utilization.

Your turn: do you use any specific metrics in monitoring your site performance? What are the biggest performance issue you face in when it comes to your ecommerce site?