Google's weighting of Page Load speed

[tdsnet]

Hey, we recently optimised our website page load speed as part of our overall optimisation.

Page load speed according to the Google Developers Page Speed test was previously 51 out of 100 and is now 92 out of 100 and this was improved within the last seven days gradually, cached half way through the improvement process.

I appreciate this is regarded as only a small part of the whole process, however, I’d be interested to know if anyone has a concrete opinion/proof on whether such a big improvement would actually make a difference to our rankings in the SERPs.

[tdsnet]

Thanks Alex, nice share.

[Alex-Harford]

Matt Cutts stated that only in 1 out of 100 searches does page speed show a noticeable change in rankings, and there are only 1 in a 1000 sites which have slow speed as a big issue: www.youtube.com/watch?v=SO4YuDAkplU

I have a page on one website which has on-page SEO close to perfection for a relatively competitive term, and a few relevant links coming in. There was a query in the SQL on the page that slowed it down - I think the load time was about 4 or 5 seconds. After a few weeks (it wasn't an important page and I knew it would be an interesting test) I fixed the problem and got the load time down to around a second. The following day the page jumped from nowhere (not even the top 20 pages) to page 4 on Google, so that must have been one of the 1 in a 100.

[tdsnet]

Thanks, some interesting stats. Our main pages have a pretty good bounce rate but the blogs are not as good. I think we may find an advantage here.

[YannickVeys]

I think, as Barry says, you'll see more improvement in what your visitors do on your site, then what Google brings you. Don't mix up cause and effect. I think because your visitors stay longer, Google will see that. And reward you accordingly.

The numbers don't say a lot, it's more the actual load time. You'll see that bounce rates and pages per visit will go up. I manage a website where we managed to cut the bounce rate in half. avg visit time up 40%, pages p/visit up 30%. Our visit rate went up by nearly 80%. So that was remarkable. And guess where the visitors came from: Google.

[tdsnet]

Thanks for the reply, sat at position 2 before the change and still sat at position 2 now after the change but hey expect we would be there so no great surprise.

Checked it in Pingdom too as we had researched this before making the changes. Interestingly I ran it 5 mins ago and it said load time 4.3 secs. Ran it again and it comes up as 1.8 secs. Seems a bit buggy today.

My opinion is that it wont make a significant difference on it's own but as part of an overall aim, it makes a slight difference.

Just wondering if anyone has seen significant improvements and can demonstrate them.

[StalkerB]

Well, surely you're now in an excellent position to tell all of us?!

I don't think the issue is whether going from 51 > 92 will help, but more whether you've gone from X seconds to less than 1.5 seconds.

In webmaster tools if you look at labs > site performance, you should see a graph of your time.

Can also use something like - http://tools.pingdom.com/ - to check how fast a page loads.

I would suggest collecting your own data (especially since you've already made the changes) and evaluating whether it's been worth it or not. You should also monitor things like conversions and bounce rate.