Announcing that page speed is important for SEO would not be a major breakthrough.

Ever since Google declared that page speed is one of the factors used to determine the order or search results, there has been a lot of discussion amongst the webmaster and SEO communities.

I don’t want to revisit this debate and would encourage you to have a read of for an excellent summary of pretty much everything that is relevant to the debate.

What I did want to share with you was some data from Google Webmaster Tools that I noticed this morning. Whilst not 100% convincing, I think that there is some interesting evidence of the impact that page speed has on all things SEO.

The data in question is the ‘crawl stats’ report, which shows Googlebot activity over the previous 90 days:

Googlebot Crawl Stats for

Googlebot Crawl Stats for

I was looking at this report as we have been struggling a bit with page speed recently and made a mini breakthrough when changing a caching plugin. I was therefore naturally interested to see if the increase in speed (which was very clear when using tools like Pingdom) was reflected in GWT data (which I have had occasion to question in the past).

At first glance, there isn’t a huge amount to be excited by other than a marked increase in crawl activity in the latter half of February.

A closer inspection does, however, reveal some interesting patterns.

If we look at page speed compared to the number of kb downloaded, there are 3 clear spikes in crawl activity, all of which correspond to a drop in page speed:

page speed v kb downloaded

Page speed (green) v kb downloaded (red)

It is, of course, perfectly possiblet that this is a coincidence and there are other occasions where there is a drop in page speed without a corresponding increase in kb downloaded, but I cannot help but feel that the increased crawl rate is linked the drop in page speed.

It feels as though Googlebot is more inclined to crawl your site if it can see that page speed has improved.

The second comparison is between page speed and the number of pages crawled:

page speed v pages crawled

Page speed (green) v pages crawled (blue)

Again, there does appear to be a correlation between page speed increasing (i.e. going down) and the number of pages crawled (going up).

It is certainly not conclusive and the lag varies to some degree but I do feel as though there is a threshold below which Google will be happier to crawl your site and it is definitely worth trying to speed things up wherever possible.

Have rankings shot up and dropped off a cliff to mirror the crawl rates? No (although the search queries report is a sea of green in the second half of the month). To be honest, I wouldn’t expect such a rapid impact on rankings and do not believe that rankings are the only measure of success.

We are currently in the process of a bit of a spring clean on this site and I would expect to see increased crawl rates when it goes live (watch this space) but this data has reinforced my belief that page speed is important for SEO.

Increased crawl activity has to be a good thing as it shows that Google is interested in your site. It also means that any tweaks you make to it will be picked up more quickly.

Page speed is even more important for your users and this should be the most compelling reason to spend some time reducing the time it takes for your users to download pages. Nobody likes to queue and slow websites are definitely frustrating to use.

What do you think? Have you seen conclusive evidence of the effect page speed has on SEO? We would love to hear your experiences, so please comment below.