Black Friday 2011 - Web Performance Results
We monitored the web performance of the top 50 Internet Retailers website during Black Friday 2011.
With so much attention being paid to online revenue this shopping season we wanted to see how well all the major retailers handled the massive influx of shoppers.
We monitored the web performance of the top 50 Internet Retailers website (Excluding Netflix, Blockbuster, 1800 Flowers, Scholastic, and Redcats USA ) from our US Nodes using Internet Explorer 8 browser. The goal was to compare the performance of the sites and to see how well each website handled the traffic during the event.
Date: November 25 2011 from 00:00 EDT to 23:59 EDT
Interval: 5 minutes from 10 US Cities (Las Vegas, New York, Los Angeles, Chicago, Washington DC, Miami, San Francisco, San Jose, Seattle & Portland)
Metrics Calculation:Metrics display median value.
Page Load Time: The time it takes the browser to download and render the page. This is the time when the browser states “Done” in the status bar – or the mouse changes from “Waiting” state to a pointer.
Render Start Time: The time it takes for the page to start displaying something other than white space.
Render Start Time:
Fastest: Macys 399 ms
Second: JC Penney 462 ms
Third: Staples 462 ms
Slowest: Crate and Barrel 2,624 ms
Page Load Time:
Fastest: JC Penney 670 ms
Second: Gap 989 ms
Third: Saks 1,251 ms
Slowest: Target: 5,856 ms
Total Downloaded Bytes:
Lightest Page: JC Penney 161.92 kbytes
Heaviest Page: Sony Style: 3,261 kbytes
Number of Requests per Page:
Lightest Page: Dell 35 requests
Heaviest Page: Barnes & Nobles 204 requests
Availability & Downtime:
Top Sites with Worst Availability, more than 5 minutes of downtime. (Based on 1440 minutes monitoring period)

* Victoria’s Secret Uptime:
In the case of Victoria’s Secret we decided not to rely solely on our Site Availability metric (based on the response code of the web server) because their site was up all the time, however on 3 occasions they displayed a maintenance page that looked like this:
Part of our methodology was to ensure we did account for maintenance window as part of the downtime calculations. We relied on the ability to capture on top of our performance metrics the number of requests and the total downloaded bytes from our Nodes for those 50 sites. This is what the picture looked like for Victoria Secrets, as you can see their site was available from a server standpoint but not from a user’s perspective.
Victoria Secrets – Was the Site really Up?
See you on Cyber Monday! Let’s hope some of these sites learned some lessons.
The Catchpoint Team.
Photo by tshein