Web site page load speeds will become an increasingly important factor in 2010 for search advertisers, and demonstrates that page load speed performance cannot be assumed, even for some of the biggest sites in the UK.

In a sample study of 100 of the UK’s most popular Web sites, my company determined that 4% of them had page load speeds slower than the acceptable threshold set by Google, beyond which the advertiser may see increased click costs.

In June 2008, Google revealed landing page download time has an impact on a marketer’s Quality Score in Paid Search. This meant that ‘latency’ began contributing directly to a campaign’s performance and ultimately it’s ROI.

At the close of 2009 there was also much speculation over whether this would also make its way into its natural search algorithms too. Matt Cutts hinted quite heavily in that direction in a video interview for Web Pro News, where he said:

“Historically, we haven’t had to use it in our [natural] search rankings, but a lot of people within Google think that the web should be fast. It should be a good experience, and so it’s sort of fair to say that if you’re a fast site, maybe you should get a little bit of a bonus. If you really have an awfully slow site, then maybe users don’t want that as much.

I think a lot of people in 2010 are going to be thinking more about ‘how do I make my site be fast,’ how do I have it be rich without writing a bunch of custom JavaScript?’”

My company took a sample survey of 100 of the UK’s most popular Web sites across four industries – consumer electronics, clothing retail, travel and finance. These we determined from our independent sector reports which reveal the most visible Web sites in Google natural and paid search across various industry sectors, per quarter.

We tested their download speeds at the same time of the day, outside of seasonal peaks in server load and made multiple requests to get an average. Our survey found load times ranged from the exceptional (Argos.co.uk at 0.29 seconds was the standout) to the painfully high (a top high street electronics retailer at over 15 seconds), and everything in between.

To define what would constitute an unacceptable average load time, we devised a methodology that mirrored Google’s method of determining an acceptable maximum. The threshold we determined was 4.97 seconds (i.e. 3 seconds above the national average). Anything above this would almost certainly fall foul of Google’s Quality Score download time guidelines, as outlined here.

Results revealed that:

• 4% of the sites analysed had average page load speeds far in excess of the 4.97 seconds determined as the threshold and therefore run a risk of seeing increased costs per click
• 3% of the sites had average page load speeds in excess of 8 seconds, which research points to as being the point at which users are likely to abandon a site
• The best performing sites, in descending order were Argos, River Island, Holiday.co.uk, Fool.co.uk, and Comet, all of which were exceptional within the group
• There was no industry pattern – all sectors had a broad spread of high and low page load speeds and the size of the company made no difference either

Google, incidentally, if it were part of this study would have performed best of all the sites as it had an average page load speed of 0.11 seconds – definitely leading by example.

Approximately 4% of the UK’s most successful websites have page load speeds that are to the detriment of their Paid Search Quality Scores. This affects their Paid Search performance and will be compounded further if Google decides to use latency in its natural search algorithms too.

Ironically, poor download speed is actually very easily fixed. Businesses need to look at things like content distribution, cache control, and even simply reducing the number of HTTP requests their pages make.

Related Articles