I came across Website Grader via Guy Kawasaki’s website today and I have to say this is one of the worst ideas I have ever seen, worse than that it’s downright dangerous in the wrong hands!
Ok let’s get the easy stuff out of the way, am I mad because I scored low or something like that … hardly this blog got a 98. Am I taking a swipe at a competing product that I or one of my friends is offering … nope. This is a bad product because it tricks people into thinking things are “OK” when in reality they aren’t (hey Jason Calacanis if you want to start talking smack about SEO being silicon snake oil this is where you should looking).
Let’s take a look at the top blog on their report engadget.com which scored a possible 98 out of 100. That’s a pretty good score so there really shouldn’t be too much wrong with that right … (leans back and cracks knuckles).
1) Engadget has a wildcard URL problem so I can link to them like this and it works http://wedontknowhowtoconfigureawebservercorrectly.engadget.com/2007/10/04/new-york-city-taxi-drivers-threaten-to-strike-again/
I wrote about this back in May and nobody cared then and nobody cares now.
2) try to pull up a bogus URL on engadget like this
http://www.engadget.com/2007/10/04/we-still-dont-know-how-to-set-up-a-webserver/
Not only do you get a page that doesn’t tell you that you went to a bad URL, but it returns a 200 status code instead of a 404 see for yourself.
3) A machine is never going to be able to give you guidance on how to focus your internal anchor text, for example here’s a post from the homepage
They are using words like [chatter] and [street] as anchor text which really helps no one. Ok we’re all lazy and do it from time to time, but a program is never going to be able to distinguish between an occasional minor lack of focus on internal anchor text or a systemic wide failure. Don’t think stuff like this matters look at the New York Times.
4) According to Google’s TOS you aren’t supposed push search pages into the index but engadget has 900+ search pages indexed. Is it a horrible offense no, but it’s an easily correctible one.
5) Robots.txt file never had one never will, but hey at least they are serving up a 404 header code this time 
Five things that could be fixed pretty easily and I only spent 15 minutes looking. Ok heading off a question at the pass now. Gee Gray if SEO matters so much how come Engadget has like a billion page views a day and buckets full of money? Well they have something called defensible traffic, built in subscribers who come to them everyday without using Google to do it. Now just think of how many more people and how much more traffic they could get if they did pay attention to SEO …
Back to Website Grader, here’s my problem in a nutshell, there are somethings in SEO that you can automate, there are some things that you can’t and a site audit is at the top of that list. You can use tools to help you gather that data, sift through it, sort, arrange it, and dissect it anyway you want. At the end of the day you need a an experienced analyst interpreting that data for you. You don’t want the SEO equivalent of a check engine light. There’s reason site clinics are so popular at search engine conferences, people need to expert advice on how to fix things from time to time.
So let’s play worst case scenario, someone with a website who just discovered SEO finds website grader runs it on their website and gets a “good score”. They figure their website is good enough and leaves things alone, chances are that person won’t have the luxury of defensible traffic that engadget does, and could really use the extra traffic that and SEO can bring to the party …
Lastly I’ll leave you with this quote
“Computers are incredibly fast, accurate and stupid; humans are incredibly slow, inaccurate and brilliant; together they are powerful beyond imagination.” — Albert Einstein
Popularity: 34% [?]
Related posts:
- How To Figure Out What Parts of Your Website Aren’t Being Crawled When Googl
- New Website ViralConversations.com So if you
- Yes Virginia You Can Hurt Yourself With Duplicate Content So last we
{ 1 trackback }
{ 9 comments }
Disclaimer: I’m the developer of http://www.WebsiteGrader.com.
You make some interesting points, but not sure you sufficiently support the thesis as to why WebsiteGrader is a “bad idea”.
1. I agree that expert humans are almost always better than automated tools at things like a website review.
2. You demonstrated this deftly with the Engadget example. Some of the things you found wrong with the Engadget site actually could have been automated (and I’m adding it to the future features list), as an expert, you’re always going to find things that are not automatable.
3. I disagree that there’s a high risk of WebsiteGrader misleading users into a false sense of security. By definition, about 80% of the sites graded will get a score below 80. Those that get above an 80 likely got there by knowing a thing or two and are less likely to fall into a sense of complacency.
4. WebsiteGrader does not claim to find “all” problems (this would be a false claim). It looks for specific things and reports on specific recommendations.
5. I disagree that the product “tricks” people into anything. It’s a tool. It actually fights *against* the snakeoil SEO salesperson by actually increasing transparency and providing something reasonably simple that mere mortals can understand.
In any case, thanks for looking at the tool. It’s always good to get opinions from expert users out there. I’m disappointed you think WebsiteGrader is a bad idea, but one of the great things about the web is that it supports conflicting opinion.
If we do end up adding some of the features you’ve indirectly suggested by way of the Engadget example, I’ll let you know and buy you a beer.
I really like the Einstein quote too. Hadn’t seen it in a while.
Cheers,
Dharmesh
Inventor – WebsiteGrader.com
Meant to say “Inventor” in the signature. Have not had my morning coffee yet.
Dharmesh
*inventor*, WebsiteGrader.com
this tool works a lot like the seomoz page strength tool, but i like this one a lot better because
1. the logo/navigation graphics are sick
2. constructive criticism is offered in the report
3. readability grade school level is calculated
i did notice that the del.icio.us bookmark count didn’t work for my site, though. google finds 5 when i do a site:del.icio.us sitename
@Dharmesh I fixed that for you
with the page strength tool it’s pretty clear it’s just a tool for looking at a single page in isolation. With website grader the idea coming in is that this is more of an SEO checkup. You and I may know it’s not all inclusive, but “man on the street” so to speak isn’t going to get that this is a quick glossing over of some common technical aspects.
Don’t market it as tool to replace a site audit, instead market is tool used to help with a site audit.
Michael,
I couldn’t help but write somewhat of a “rebuttal”.
Here it is:
Why Free SEO Tools Bring Power To The People
Should generate some interesting commentary from both sides of the argument.
And by the way, I owe you a beer. I’m adding the ability to detect erroneous “not found” pages returning a 200 status code in the next version. That’s a good idea.
Cheers,
Dharmesh
Hmmm… why has no one pointed out the obvious data mining that this “free” SEO tool provides for the owner? The owner gets to know that a site is interested in SEO, what that site owner considers to be primary keywords, who the main competitors are, and an email address is required. Of course this is lead generation for SEO services, SEO software, keyword research, etc etc etc.
Check the privacy policy… it says quite clearly your email address and everything else known about you will be used to market to you (including affiliate products).
Argue all you like about how good or bad it is, but please stop saying it is “free” for the good of the users.
I’m inclined to defend Websitegrader as a measure of internet marketing effectiveness – which is a slightly different animal from SEO. I view SEO as an important part of an effective internet marketing campaign – not the whole enchilada.
When Darmesh did his entry on the top blogs, the thrust was about how bad the SEO was on some of those blogs. The fact that those blogs still scored very highly on the WebsiteGrader despite hopeless SEO shows the tool works as it’s supposed to.
Great post, Michael. I’m currently in the process of systematising and reengineering my SEO service offering so I’m really interested in this topic right now.
I agree that much of an SEO site audit cannot be automated cost effectively. However, I also believe that with a big enough budget to develop tools, all of the issues you raised there, and a whole host that you didn’t mention, can all be automated.
There is a very large but finite number of issues that one can find during a site audit. It might even be possible to develop tools to find 80%, or 90%, or 95%, or even “Five 9s” levels of problems without human intervention, at an affordable price point.
Right now I’m not interested in developing tools to automate to that level, but from a business perspective I certainly am interested in using existing tools to make both audit and reporting as efficient as possible.
- Alastair.
I like the Website Grader but… I don’t think it should be taken all too seriously. It can help fix specific problems and that’s good. But the grade itself isn’t that meaningful, I don’t think.
Comments on this entry are closed.