Welcome to a new monthly column where we look at Web performance metrics for the leading US lodging and travel agency sites. Keynote measures web performance 24/7, reporting every week on how these global brands rank in our Index.
Each month we will go behind these numbers to take a closer look at what is affecting particular sites’ performance in either positive or negative ways. In the process, we will share with you some thoughts around best (and worst) practices for delivering the optimum web performance for visitors.
Measuring the True End User Experience
We measure home page performance, which means how fast a visitor gets the page to display. Most benchmarking of page performance starts with the load time, or response time as it is commonly referred. New industry standard timers supported in browsers can also now reveal performance more holistically—based on users’ true experience during the page load. Website owners can now easily measure the various events in the process of navigating to a page, including timestamps for the starting and ending of key phases:
These are important events in a page’s lifecycle and make the difference between a user seeing something – and seeing nothing. As an example, imagine you visit two sites, each with a page load time of five seconds. If site A starts rendering content in half a second and site B doesn’t start showing you any content at all until four seconds, your perception of site A will be much more positive even though both sites have the same overall 5 second user experience time.
The first point at which the user sees something other than a blank screen (“Time to First Paint”)
When a page can be fully clicked, swiped and scrolled (“Time to Interactive Page”)
The total time a page takes to render completely (“Total User Experience Time”)
So let’s take a closer look at this month’s Keynote Lodging Performance Index. It measures and benchmarks the performance of the home pages of the major hotel and travel booking sites from the ten largest US metropolitan areas (Boston, Chicago, Dallas, Detroit, Houston, Los Angeles, New York, Philadelphia, San Francisco and Washington, DC) on high-speed links attached to key points on the largest US Internet Service Provider (ISP) backbones. Sites are measured every fifteen minutes.
Dislodged Lodging Sites
For our first analysis we decided to focus only on the lodging sites on the index excluding the aggregators and travel agency sites.
For the period between May 1st and 31st, Hyatt rules the Index – except for one day when Best Western sneaks into first place. Otherwise it’s Hyatt in number one and Best Western in number two position throughout. Clearly both are well-optimized sites.
But here’s the interesting observation – we see Harrah’s go from being consistently near the top ranking, to plunging to almost bottom (Hilton usually has a lock on the bottom position). What happened?
A standard industry benchmark for load time (not just for lodging but for all consumer-facing sites) is under four seconds. Even four seconds pushes many of our patience. But on the Index we see Sheraton, Harrah’s, Wyndham Worldwide and Hilton all going over that threshold.
Looking closely at the Harrah’s site we see that a page redesign around the middle of the month jacked load time from approximately 1.8 seconds to 7.5 seconds. This redesign also increased the page size from approximately one megabyte to more like three megabytes, which suggests the redesign was intended to push more rich content. But did the redesign backfire?
Time to First Paint is a respectable .9 seconds. However, the problem starts as the Harrah’s site struggles to load a large number of images. This dramatically impacts the “Time to Interactive Page” to a poor 5.4+ seconds. Many Web sites want to deliver a rich content experience for the user – such as sleek graphics and video - especially sites in the lodging and travel industries. However, rich content is usually where performance problems start.
This is where Harrah’s is suffering – lots of icons which could be much better managed in CSS Sprite files. Using CSS sprites will reduce the overall number of elements downloaded by the browser saving the time it takes to go back and forth between the browser and the server. The best practice recommendation is to load all CSS as soon as possible after the base HTML document for the web page. It sounds simple, but merely changing the order in which assets are referenced in the HTML document can dramatically change the user experience of the page load. Site owners should also track initial render (Time to First Paint) performance as a key user experience metric. Once the changes are rolled into production, continue to monitor the initial render experience with a program of ongoing site monitoring.
We also see that Harrah’s is not using a Content Delivery Network (CDN) such as Akamai. This is unusual as CDNs play an important role in hosting content locally to a user to avoid the reliance on one data center and host the site’s content as close to the users as possible. Right now we see San Francisco waiting nine seconds and New York almost five seconds for a response. If a site chooses not to use a CDN then a major requirement is to limit the number of roundtrip requests to the server.
Harrah’s is a good example on how a change in site design can bring about major changes in Web site performance. To minimize this impact, consolidation of images to reduce the overall number of elements or adoption of a CDN by Harrah’s would help optimize the user experience of their home page.
Anyone can also sign up for a free monthly email delivery of the Index. Use it to track how your company’s performance is doing against the competition, or just to follow what some of the major names are setting as performance standards. Keynote runs a large number of US and global Indexes, across a range of industries and government, which many organizations use as the benchmark to achieve their own optimum Web performance.