« | Home | »

Bullshit Video: Google Site Performance for Webmasters

By SEO Mofo | May 4, 2010

I hereby declare the Google Webmaster Central Blog’s latest video post, Site Performance For Webmasters, a bunch of bullshit. Without exaggeration…this is absolutely the worst Google video I’ve ever seen. The video contains all kinds of bullshit, but I’ll just mention the things that really annoyed me. Go watch the video, then come back and listen to me bitch about it.

Need for Speed

The first part of the video addresses the question: “Why does Google think speed is important?” Only it doesn’t. Instead, the video makes the following 3 points:

The problem is…that answers the question: “Why should webmasters make their sites faster?” That has nothing to do with Google. People don’t like to wait–we know that already. We want to know why Google is telling us to make our sites faster. What’s in it for you, Google?

Faster on the frontend: for little or no money down!

This is incredible…look at the first slide from this part of the video. The video references the work of Steve Souderswho is an employee at Google–and yet they misspelled his name?

Bullshit Video: Google Site Performance for Webmasters

Available tools

First of all…if you’re trying to make a video that urges “beginner to intermediate” webmasters to take the first steps toward speeding up their sites, then maybe you should use tools and software that those webmasters are most likely to be familiar with. For example…let’s say you want to show webmasters how easy it is to throw a few caching directives into their .htaccess file. Personally, I’d show them how to do it using cPanel or FileZilla–you know, something a lot of people can relate to.

What I wouldn’t do…is show how to edit an .htaccess file using an SSH terminal and a command line interface. And if I did use an SSH terminal and a command line interface, then I’d at least try to copy and paste the .htaccess directives without inexplicably dropping a line of code and making the entire thing useless…

Bullshit Video: Google Site Performance for Webmasters

Bullshit Video: Google Site Performance for Webmasters

FAQs

What’s a good response time to aim for?

Google’s answer: according to one study (that Google didn’t conduct), for e-commerce sites…aim for 2 seconds. Google aims for under half a second.

In other words: if you own an e-commerce site, then aim for the totally-unrealistic-and-arbitrarily-chosen time of 2 seconds. If you are part of the team of Google engineers that works on making Google’s search results appear faster, then aim for under half a second. Everybody else…go fuck yourself.

Does progressive rendering help users?

Google’s answer: Yes

In other words: yes, progressive rendering makes users think your site is faster…but Google has absolutely no way of knowing whether or not your site uses progressive rendering, so it will NOT help your search engine rankings…at least not until our overzealous engineers have sufficient time to write an entire algorithm to compute progressive rendering…and even then, it wouldn’t work…because our engineers are narrow-minded, antisocial nerds and no one checks their work.

How can you implement progressive rendering techniques on your site?

Google’s answer: put stylesheets at the top of the page. This allows a browser to start displaying content ASAP.

In other words: I have no fucking idea. Go buy Steve Soundgarden’s books and stop bothering me.

3 Steps to Success

  1. Check out Site Performance in Webmaster Tools
  2. Install Page Speed
  3. Explore! Check out tools like YSlow, WebPagetest.org, hang out in the “Make the Web Faster” forum

I’m confused…how do those steps lead to success? Is knowing your site is slow the ultimate goal? Are you saying identifying the problem IS the solution?

Site Performance is bullshit, Page Speed and YSlow are virtually identical, and no one knows WTF this forum is. The only decent recommendation is WebPagetest.org. Other than that…this entire section of the video is useless bullshit…and it’s nothing more than a rehash of the material from the “Available tools” section.

Looking ahead: Performance and SEO

Check this out…right when the video gets to the part all of us SEOs wanna see…it violently erupts into a thermonuclear meltdown of astronomical bullshit. The first piece of bullshit can be found in the following slide. Notice how the “Measurable SEO value” is the result of several factors that have absolutely nothing to do with SEO. Honestly…I’m 90% sure that the person who created that slide doesn’t know what SEO stands for. Look at this bullshit:

Bullshit Video: Google Site Performance for Webmasters

If by some miracle you’re not completely confused by that bullshit slide, then surely the following quote will leave you scratching your head:

Let’s summarize by taking a look ahead at how performance can actually impact the SEO industry.

If a faster site has now been proven to increase conversions, page views, and time on site, meanwhile lowering bounce rate and operating costs…

In fact, Shopzilla found that by just speeding up their site, it decreased their operating cost by 50%.

You like that? It’s like a season finale cliff-hanger on your favorite TV show…but then the following season opens with a new cast and a new location, and the show never resolves the cliff-hanger from last season. Good times.

Summary

In my opinion, this video is a sobering reminder of how little Google monitors the quality of information they publish. The quality of this video is comparable to something I would expect from a college football player delivering a speech in front of his Public Speaking class…on a Monday morning…after a long weekend of partying. I find it absolutely disturbing that Google can hype up website speed as much as it has…and at the peak of the hysteria, they throw this garbage at us?

Clearly, Google needs to overhaul their internal structure to incorporate a better system of checks and balances. Their current system of autonomous teams continues to fail, because it produces results that are only as good as the team’s strongest link. Google needs to increase communication between these groups to ensure that everything Google publishes or releases to the public has been double-checked, triple-checked, etc. Too much isolation and independence can be harmful to a company’s brand, as we’ve already seen with Google Buzz. For a company as big (and wealthy) as Google, it’s very difficult to justify mistakes that would have easily been prevented by an additional layer of quality control. Google needs to stop measuring success by the rate at which they produce stuff, and start focusing more on the quality of their current services.

/rant

Oh wait…I almost forgot to link gratuitously to my own website. So…um…be sure to check out these informative posts about page speed, learn how to make a custom Twitter background, use this free tool to optimize your Google SERP snippets, and download free middle finger cursors. Peace, I’m outta here. -SEO Mofo

No Bullshit? Then just Share it!
  • Bullshit Video: Google Site Performance for Webmasters

Topics: Other Stuff | 25 Comments »

  • http://sebastians-pamphlets.com/ Sebastian

    You have spoiled me the fun. Can't view the video unbiased now. Should have taken your advice to consume it upfront.

  • http://yoyoseo.com DanaLookadoo

    Darren!!!!!!! WELL done analysis about Google's lack of editorial analysis!

    Peace

  • http://seriocomic.com seriocomic

    Who cares? Maile is hawt.

    Seriously though, I thought I had attention to detail but didn't pick up on Souder's name being mis-spelt. I did think the use of SSH to edit the .htaccess was a WTF moment.

  • http://www.thelostagency.com/ David Iwanow

    Hahaha good post, if you are bored do some load tests on various sites starting with http://adwords.google.com and see if your site performs better, Adwords is around 4 times slower to load than most of the Google properties including Youtube…

  • http://www.navinpoeranc.om Navin Poeran

    Maybe I don't understand what you mean, SEOmofo.

    But how are “conversions, time on site and pageviews” unrelated to SEO?

  • http://www.making8seo.com/ Adam J. Humphreys

    Unfortunately there's been a lot of bugs coming from Google lately that go beyond sloppy such as Google Buzz, Google Maps, and unfortunately some of the information in this video. I've found a lot of the content in their general documentation for webmasters to be not only dated but often explicity vague. While I can appreciate that Google's an awesome company I sure hope they can fix these issues. It seems kind of embaressing that a company of this scale could make such large changes without testing them for bugs. Matt Cutts recently mentioned in a video that with the Google search algorithm alone they make 350-400 changes a year. One of the changes to Google Analytics (Conversion University) I hope they add is a mic filter (bleeding ears all around).

  • http://zemalf.com/ Antti Kokkonen

    Here's what I learned: viewing a video, knowing the stuff in it beforehand, makes me ignore all the mistakes in it – thanks for writing them out! This analysis made the video much more fun, less good, less useful to anyone, but fun. Leaving the ExpiresActive On away from the .htaccess is an instant classic (if using SSH didn't make it one already).

  • http://www.propdata.net Robert

    Fantastic overview. Good long rant, but one line really sums it up:

    “I would expect from a college football player delivering a speech in front of his Public Speaking class…on a Monday morning…after a long weekend of partying.”

    Say no more! :)

  • http://www.metayer.de Jack

    Ok you don't agree on some points and the video could give more usefull hints, but: it gives Webmasters an idea where to start and tell them to spend some time load time.

  • chapter42

    I agree with this piece, allthough I also strongly believe in making your sites faster. For instance this Mod_expires ins't installed a lot since it's quite buggy and mostly not installed by default. Maybe how to compile this thing for your apache should be a nice video as well.

    Maybe even more funny is that most of Google Javascripts are even the worst snippets to put on your speedwise.

  • Tym @ PPC_Management

    LMAO! Yep, just look through ANY of Google's *documentation* and it's like a 3rd Grade idiot wrote it. I guess all the PHD's are working on the algo. :)

  • http://www.SEOmofo.com/ SEO Mofo

    Okay, so I exaggerated a bit. Certainly in SOME contexts, SEO can be related to everything on that slide in one way or another. But in the context of this video, it made absolutely no sense. If SEO meant something like “the success of a website,” then the slide would make sense.

    As it stands, the slide suggests that increasing conversions, time on site, and page views, while decreasing bounce rate and operating costs, will result in “measurable SEO value.” That's just not true. At a bare minimum, the word “measurable” would have to be removed before anyone could even try to make a case supporting that statement.

  • PokerSEO

    IT's honestly very nice to see people finally start to question and call out not only Google but other “accepted” “Keys” that are always spouted on forums, blogs, etc.

    According to google's little tools, my sites are sorta slow, yet they seem to convert really well and no human seems to think they are slow. So what is all the fuss really about?

  • http://johnnyblog.info Johnny

    You're right, you're absolutely right – this was a really sloppy job that Google should be ashamed of publishing.

    I do have to say this though (and I don't want to detract from that eloquent tear down above, seriously I don't) – the lady tells us flat out “because speed is so important, and because we aim to give users the best search experience possible, site performance is now a factor in google rankings.”

    I think that what they really wanted to say (but failed) was that because increased speed equates to the possibility of increased visitor satisfaction, and if nothing else averts certain types of dissatisfaction for visitors, and because Google's whole motto is tayloring their search results to give users the highest level of satisfaction, they are now including it as a factor for their search rankings, thereby making it a new form of search engine optimization (because by optimizing for speed, you'll possibly increase your ranking).

    They probably could have said that then spent the other 11 minutes doing a decent job of everything they screwed up that you've pointed out! XD

    Thanks so much for that post! It was really great!

  • Ronabop

    cPanel? Filezilla? Seriously?

    Why not insist on having directions for editing Apache directives using FrontPage, or MS Word, or another method of “Internet: now with training wheels for mouth-breathers!”?

    Oh, wait, that's exactly what was done.

    As far as the laughably slow two second metric, yeah… a bunch of non-programmers may find it hard to optimize to sub-second speeds. Which is why they aren't programmers, and are stuck with bitching about using other people's crapware that takes ages to do something simple.

  • http://www.SEOmofo.com/ SEO Mofo

    @Ronabop,

    If your programming logic is anything like that of your blog comments, you need to reconsider career paths.

    Your first point is invalid because you failed to realize that this video is in fact published for the benefit of “mouth-breathers.” Nose-breathers get their information from books and expert sources, not from Google videos.

    Your second point is invalid because you failed to understand WTF we're talking about in the first place. The time it takes to load a page into a web browser is largely dependent on network latency, which cannot be avoided, regardless of how well the page/server are programmed. I'm sure you and your badass Dungeons & Dragons crew can load each other's plain-text attack stats lightning-quick over your local network, during one of your infamously wild LAN parties, but that's irrelevant to the topic of this post.

  • http://www.huomah.com theGypsy

    lol… another keyboard ruined… thanks bro.

  • http://twitter.com/kevinmspence Kevin

    Two keyboards, one comment.

    Thanks, mofo.

  • sabsa

    So are they pushing for that half a second so they can take the remaining 1,5 to load the google-analytics.com service?

  • http://www.metayer.de Jack

    @sabsa there was an analytics code update about seven days ago. To avoid delays you have to insert a new code in your page.

  • Jackoh

    This is such bullshit. “Largely dependent on network latency”? You really have no idea what you're talking about. It's a factor, yes, but a bigger factor is not serving static content from a domain with cookies, or enabling gzip compression. Faster latency will only help so far is both those situations, but they can both double your page loading time.

  • Navin Poeran

    lol

  • http://www.SEOmofo.com/ SEO Mofo

    @Jackoh,

    You missed the point entirely. Ronabop is claiming that page load times under 1 second are easily attainable to savvy programmers. I'm saying it's virtually impossible to load an entire non-cached web page in under a second, because network latency (i.e., DNS look-ups, HTTP requests) cannot be avoided. I never said it was the primary cause of slow page loads.

    If you are interested in participating in public forums, I recommend you first bring your reading comprehension up to at least a 5th grade level.

  • Mark

    Maybe their crawl costs will drop because the first thing most site owners will do is half the size of their images. Its all about google google google – dont ever forget that.

  • Gina Bisaillon

    I stopped watching Google’s videos after coming to the conclusion that they leave it up to any kid who works for them to decide what content to shoot, how to shoot it, what to say, etc. They’re so amateurish it’s pathetic! For one of them, the sound was so bad that all the comments were about the fact that they couldn’t make out what the googler was saying. If Google won’t invest in professionals for that, it makes one wonder about the rest of their operation.