Just my two-cents worth.

At the Pubcon conference in Las Vegas (Nov ‘09) you are reported to have indicated that good loading times may positively influence a site’s SERPs (though the opposite does not hold true).

Source: http://www.seomoz.org/blog/new-interesting-insights-into-google-rankings-spam-from-pubcon – second bullet point.

It would be great if you could confirm that here for anyone stumbling over an olb blog entry and mistaking it for current advice.

Cheers

I would rank fast sites over slow sites for the sake of good user exeprience to the Google users. Though the ranking nuances might be slim but there it should be in my opinion.

So are you saying, site A that’s constantly 5 times slower than site B will be no inferior to site B in Google search rankings at all?

With due respect, I don’t know if you have close contact to the technical staff of Google (or you are one of them?), or maybe you are just busting the myth by taking the myth automatically incorrect, please if you can, can you confirm this with someone who’s directly responsible for the ranking algorithms.

Does the google bot parse just the HTML or actually load an image? Meaning, if my HTML page size is 50KB vs the actual page size with all the images and flash and what not is like 500KB, does the google bot load 500KB like a browser or just the 50KB?

And how will this affect indexing?

Appreciate a response.
Cheers.

Someone is doing this on purpose. Got the time to investigate? If my investigation has turned up the right person, I have no choice but to contact authortities, unless you can assure me in some fashion that he isn’t harming my business……it seems as though he is…..late September, all msn traffic gone…..a few weeks later, all yahoo traffic gone. If all google traffic disappears, his name makes headlines.

I’ve been in contact with a .edu domain administrator who is willing to help since fake p*rn links to other sites were found within a wiki ( wiki.cs.uiuc.edu ) on their .edu domain. They’ve altered the page ( wiki.cs.uiuc.edu/CISTUG/&username%3Djameno123 ) since I notified them, after a google search, about this person. I think he lives in Amelia, OH….

For instance, with curl, I can do:

% curl –output /dev/null –write-out “Total time: %{time_total} s, DNS time %{time_namelookup} s, connect time %{time_connect} s, %{size_download} bytes transferred\n” http://www.mattcutts.com/

% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 471 100 471 0 0 2382 0 –:–:– –:–:– –:–:– 0
Total time: 0.198 s, DNS time 0.003 s, connect time 0.004 s, 471 bytes transferred

And see that the network is fine and the download time reasonable.

curl did not get images, so if images are an important part of the page, you may have to use another tricks (left as an exercice for the reader).

Also, if you consider user experience and not only network time, very often, the time to display depends, not on the lot time, but of the display time (Javascript to run, complicated CSS, layout, etc).

Disclaimer: I’m the author of echoping so the moderator is free to regard this as unsollicited advertisment