Improve Page Speed Score by not displaying ads to googlebot?

[aish1108]

Page Speed in now a Search ranking factor.

Here's the problem. I can't get below 4.5 seconds on GWT because on every page of my site there are 3 ads from Google Adsense and other Ad networks.

Page speed is measured both by the Google Toolbar and Googlebot. Can I just not show ads to Googlebot?

2 questions
1. is this considered cloaking?
2. Will it make a difference in my page speed score, since the Google toolbar is also used to measure page speed

[topr8]

>>Can I just not show ads to Googlebot?

there is nothing wrong with not showing ads to bots.

but in this case ... adsense is an external javascript call and i'm not sure that googlebot calls external javascripts, so it won't affect anything.

[eltercerhombre]

The problem with Google's way to measure speed it that is using Toolbar data, which makes pages THAT ONLY I USE to look my site like slow, when is faster than the 90% of the websites.

[tedster]

Page speed is measured both by the Google Toolbar and Googlebot.

I've never seen anything from Google that mentions googlebot speed data being used for the Site Performance ranking metric - have you go a reference for it?

[rustybrick]

tedster, you are right, Google uses toolbar data for this, see [google.com...]

"It is collected directly from users who have installed the Google Toolbar and have enabled the optional PageRank feature."

[aish1108]

I've never seen anything from Google that mentions googlebot speed data being used for the Site Performance ranking metric - have you go a reference for it?

I was relying on this article from Search Engine Land ( [searchengineland.com...] )

How Google Measures Page Speed

Singhal says there are two primary ways Google will measure page speed:

1. How a page responds to Googlebot
2. Load time as measured by the Google Toolbar

But I can't seem to find any post on any of Google's Blogs which mention Googlebot being used.....?

[TheMadScientist]

Site Performance in WMT [googlewebmastercentral.blogspot.com]

I cannot find a single resource citing Singhal and stating speed will be determined by response time to GoogleBot requests, but the resource linked above does state GoogleBot is used to make the requests for the resources to determine speed based on G's FireFox plug-in, so maybe it was a misread of that information or something that led to the statement, otherwise I don't see a source for the statement or resource linked to ensure the statement is correct and the Google resources seem to contradict it...

[edited by: TheMadScientist at 9:16 pm (utc) on Apr 13, 2010]

[tedster]

Thanks. Well, we do know that Google keeps a record of server performance with regard to googlebot. The question here is whether that particular data will be part of the ranking algo change or not. It's possible that Amit Singhal's remarks have been slightly "lost in translation" in this case. We'll try for some clarification.

At any rate, I don't intend to do server yoga just to make site performance look better to googlebot - especially since Google will have toolbar data from users anyway. There are SOOOO many other ways to get major boosts in site speed.

Check out those two main speed tools [webmasterworld.com]. The research behind them uncovered some best practices that were definitely not part of common knowledge before Steve Souders dug in and tested for real.

[TheMadScientist]

The worst thing is if Search Engine Land is incorrect we'll be trying to put the myth to rest for years because of a single trusted resource's mis-statement or mis-understanding of the approach G is using...

[aish1108]

There are SOOOO many other ways to get major boosts in site speed. Check out those two main speed tools [webmasterworld.com].

True, but as anyone with a website serving ads from multiple Ad Networks will tell you, the Page Speed Tools and therefore, I assume Google, don't differentiate between content being served from your web server and Ads being served from Third Party Ad Networks such as ValueClick/fastclick


For more on ValueClick/fastclick see Steve Souders post from today [stevesouders.com...]

Grade F on Use a Content Delivery Network (CDN)

There are 50 static components that are not on CDN.

You can specify CDN hostnames in your preferences. See YSlow FAQ for details.

* http:// www.google-analytics.com/ga.js
* http:// www.google.com/coop/cse/brand?...
* http:// partner.googleadservices.com/gampad/google_service.js
* http:// partner.googleadservices.com/gampad/google_ads.js
* http:// as.casalemedia.com/s?...
* http:// www.google.com/cse/intl/en/images/google_custom_search_watermark.gif
* http:// tags.expo9.exponential.com/tags/12345/ROS/tags.js
* http:// a.tribalfusion.com/j.ad?...
* http:// media.fastclick.net/w/get.media?...
* http:// s0.2mdn.net/879366/flashwrite_1_2.js
* http:// amch.questionmarket.com/adscgen/st.php?...
* http:// s0.2mdn.net/1649785/Motorola_Devour_SmartSleekSocial728x90.swf?...
* http:// t.mookie1.com/t/v1/imp?...
* http:// a.tribalfusion.com/j.ad?...
* http:// media.fastclick.net/w/get.media?...
* http:// ad.doubleclick.net/.../click.here?...
* http:// s0.2mdn.net/961941/160x600_rumor.swf
* http:// a.tribalfusion.com/j.ad?...
* http:// media.fastclick.net/w/get.media?...
* http:// cdn.fastclick.net/fastclick.net/cid51376/v8flash.js
* http:// cdn.fastclick.net/fastclick.net/cid220122/38380_Frogger_300x250.swf?...
* http:// cdn.fastclick.net/fastclick.net/cid220122/tracknfield_160x600_0809.jpg

[TheMadScientist]

When looking at flagged issues regarding common third-party code such as website analytics scripts, one factor that can also play a role is how wide-spread these scripts are on the web. If they are common across the web, chances are that the average user's browser will have already cached the DNS lookup and the content of the script. While these scripts will still be flagged as separate DNS lookups, in practice they might not play a strong role in the actual load time.

Purely informational.
Resource linked above.

[BillyS]

It's going to be interesting to see how this plays out. A lot of us have been tweaking our sites to reduce load times. The only recommendations we have left (our scores are around 90 / 100 right now) have to do with Adsense, Analytics and we run Addthis...

[setzer]

Google's methodology behind collecting speed metrics is worrying. If your visitor base has slower internet connections then by extension your site speed will suffer.

I've employed every trick in the book to improve response times on our site AND our static content is served direct from Akamai (arguably the fastest CDN in existence). Still, with a 3.6s load time, our main domain is slower than 60% of sites out there according to the G. Bit of a bummer.

I hope they're still working out the kinks, as there's basically nothing else I can do to improve response time.

[tedster]

Welcome to the forums, setzer.

You know, Google's page speed score is related to server speed by just a small amount. If you run the Page Speed tool, most of the feedback you get is all about on-page factors, coding factors of many kinds. This is because Google wants to measure how long it takes before the page APPEARS on the average user's screen, not just how long it takes for the server to send back the code.

I've been doing site speed metrics and optimization for my clients for seven years, and I've been attentive to this area from back in the mid nineties. I always considered site speed to be one of the secret weapons for a successful site.

It's been an open secret because a good bit of the information has always been available but so many websites didn't seem to care -- at least not before Google said it was going to be a (small) part of the ranking algorithm.

In all that time, I never saw a site that couldn't improve their speed without even touching the server. I interacted with many site builders who were reluctant to do the work, but the speed improvements are almost always there to be done. And with the recent insights from Steve Souders' research [video.yahoo.com], even more opportunities exist.

And by the way, Steve now works for Google ;)

[curioustoddler]

Google wants to measure how long it takes before the page APPEARS on the average user's screen

Very well said. Google will look at it from user's point of view. So size and number of images, html code, your server side coding etc. will be more important factors.