10 replies to this topic

[cre8pc]

FYI

Web Sites That Suck Increase Stress

We know that slow, balky, and confusing websites aren’t a good thing. Traffic metrics show this, as does conversion data. Google, whom some think of as passively indexing the web, believes quick-loading pages are essential to a good user experience. Google is, in fact, actively trying to speed up websites (and keep their search users happy) by making page load time a ranking factor. (See Barry Schwartz’s article at Search Engine Land describing Google’s Matt Cutts commentary at Pubcon.)

[DonnaFontenot]

Back in November, I posted some links in the back room that would be appropriate here in this thread. I said (in reference to a discussion about whether page speed might be important to Google):

First up is the page speed tool that Google created:
http://code.google.c...eed/page-speed/

And then here's the info that talks about how page speed affects Adwords Landing Quality Scores:
http://adwords.googl...mp;answer=87144
(which of course is different, but still...)

and finally, here's Matt Cutts himself answering that very question in a video:
http://www.youtube.c...h?v=B3zmP0W26M0

(Spoiler: No, but...)

Oh, here we go...this was what was nagging at my memory...

http://www.techworld...than_revolution
and
Google's "Let's Make the Web Faster" initiative:
http://code.google.com/speed/articles/

Google has mentioned this in other ways and in other places since then, but this list is a good review of how this started being noticed/discussed.

[jonbey]

Well, I know of 2 sites that had more traffic after moving to a faster server. Not a lot of data to go on I know.

One thing we considered (I and my mate that also moved to the same server) was that it was party user driven - if people land on a responsive site they are more likely to like it and bookmark it.

But it would make sense if a faster server is featured in the search algo somewhere. It is another sign of quality. Something that can be measured too. People running hundreds, or thousands even, of MFA sites are unlikely to be investing in quality servers, whereas people running "proper" web businesses are more likely to ensure that their sites are working on good servers.

[fairclb]

That makes a sense jonbey, but could give an unfair advantage to the larger businesses while hurting the small businesses that can't afford to have redundant fiber connections with a larger load balancing server cluster. Smaller businesses could go with a shared hosting account, but in those cases you have no control over what else is hosted there and in general the host doesn't put the effort or resources into optimizing the servers and network to the level that a dedicated team with a vested interest in the success of the business would.

I like the idea of making the Internet faster for everyone, but at the same time will that curb creativity and advanced functionality that some people may be willing to wait a little longer for? It wasn't that long ago that images were unheard of on the Internet, now they are an essential part of branding and creating a sense of professionalism. There are some (Jacob Nielson) that will still argue that images shouldn't be used. Will this effort by Google take us back to the days of all text websites? There needs to be a balance here, fast is good, all text is boring.

[Ruud]

Making your site faster isn't *that* easy. Running on a dedicated server, site optimized from here to Tokyo, Google still claims 90+% other sites are faster.

If, when you make a move, armies of optimizers start to work on what you just did -- could there be any benefit at any given time to make a move in a mutually beneficial yet irrelevant direction? Could you keep people busy occupied with other stuff?

:tinfoil:

[fairclb]

Making your site faster isn't *that* easy. Running on a dedicated server, site optimized from here to Tokyo, Google still claims 90+% other sites are faster.

Most likely because that dedicated server is being compared to companies that have multiple connections to the Internet using fiber lines instead of a single T1 or similar connection. Many also create server clusters meaning they have several severs sharing the traffic load instead of a single server. Having more resources they are able to separate web servers from database servers, again reducing the amount of processing that a single machine running both the database and web server would have to deal with.

Larger businesses are also able to dedicate personnel to tuning server side caching that will allow the servers to keep frequently used information in memory rather than having to pull it from hard drives or the database for every visit.

These are just a few examples of ways that larger businesses will gain a huge advantage over equally relevant websites that have less money and resources to compete. I am all for speeding up the Internet, but should a page that loads in 1 second be considered more relevant than a site that loads in 2 seconds?

My assumption is that load times will be just one more piece of the algorithm, not the determining factor in how a site ranks. The question is, how much weight is put into page load times compared to the others factors.

[Ruud]

My assumption is that load times will be just one more piece of the algorithm, not the determining factor in how a site ranks. The question is, how much weight is put into page load times compared to the others factors.

If speed would be an issue besides number of pages crawled -- at issue is then the classic result that "When we looked at the actual download speeds of the sites we tested, we found that there was no correlation between these and the perceived speeds reported by our users."

There was still another surprising finding from our study: a strong correlation between perceived download time and whether users successfully completed their tasks on a site. There was, however, no correlation between actual download time and task success, causing us to discard our original hypothesis. It seems that, when people accomplish what they set out to do on a site, they perceive that site to be fast.

* my emphasis

If not for ranking -- what other motivation(s) might Google have to push this?

[bwelford]

That Google "Do No Evil" group are probably trying to save the planet by encouraging us all to use less energy/bandwidth. Now if only they would bury the Google Toolbar PageRank indicator that would have a really big impact on the number of spam web pages that were created.

I particularly liked a Peter Drucker phrase that was new to me this week:
There is nothing so useless as doing efficiently what should not be done at all.

[Ron Carnell]

Ten years ago, or there abouts, Google came out against popup/popunder advertising on web sites. "Popups are evil," said Google with their usual air of authority.

Google's distaste for popups never made it into the algorithm, however. This one won't, either.

If it does, you'll immediately know it because YouTube will stop ranking for anything. Those damn videos take forever to load!

[fairclb]

I have to agree with Google on the pop-ups though... they are extremely annoying.

[jonbey]

Pop ups do not affect rankings? I did not know that.