When I wrote these exam questions, I have to admit, I didn’t think about publishing the answers. It seemed kind of… wrong. But then I thought about the havoc wreaked if all kids had the teachers’ answer keys (obscure Simpsons reference there). And I smiled in a way that makes my family nervous.
So, the answers:
1. True or false or both: The ‘nofollow’ attribute is good for SEO.
Yeaaahhhh so about this one. I did a lousy job with the question, and about 30 people simultaneously pointed that out. The question should have read:
True or false: The ‘nofollow’ attribute is good for pagerank sculpting.
Then the answer is false, per Matt Cutts. ‘Nofollow’ is actually good if you’re using it to prevent penalties relating to link selling.
This is why I’m not a professor, OK?
2: When are drop down menus bad for SEO?
The answer is
c. If they cause unnecessary PageRank ‘leaks’.
And now, you’re saying “Ian, wtf?! When is a PageRank leak necessary? And who cares about PageRank anyway?!”
PageRank does still matter, even if the math behind it is now a total mystery. Somehow, every page on the web holds its own authority. Links allow that authority to drain away to other pages. Put a ton of links on the page and you reduce the amount of authority passed by each link. That means the page can no longer effectively pass authority where you want it to go. That’s bad.
The leakiest part of any page is often the drop-down menu. Somehow, drop-downs have become the place where everyone compromises: No, we won’t put the link in the drop-down. OK! OK! Stop whining! We’ll put it in the drop-down! Now leave me alone!
And you get stuff like this:
Drop downs don’t kill people. Stupid people creating bad drop downs kill people.Impossible to understand, and a PageRank sieve.
3: How many times should you repeat a key phrase on a page?
The correct answer is
a. Come off it, Ian, there’s no right answer to that.
You should use a word or phrase precisely as many times as you need to. No more. No less.
Just keep in mind that search engines have a much easier time classifying you as a puppy web site if you actually use the word ‘puppy’ somewhere.
4: What’s the most important on-page SEO element?
The answer is b: The TITLE tag. No argument. No discussion.
5: What’s the best choice for a server response code if a page doesn’t exist?
As @pageoneresults correctly pointed out, the right answer wasn’t actually in there. Given what I put in the choices, the right answer is ‘b. 404’.
But a ‘410’ response code is even better, as it results in faster removal from the index.
6: What’s the correct way to redirect…
…visiting browsers and bots from a page that is permanently gone to a replacement page?
‘a. 301’ is the correct answer. No ifs, ands or buts.
7: Under Google’s Panda update:
The right answer is:
A site with 10 great pages and 1,000 lousy ones is in trouble.
Although the bamboo thing might be worth a try.
Herrreeee panda panda panda…
8: TRUE OR FALSE: Site performance can impact search rankings on Google and Bing.
True. A slow site reduces crawl efficiency and may have a rough time getting indexed. But also, there’s strong evidence that site performance is a ‘quality’ factor, which means it goes straight to Panda. Roar.
9: Pick the most important SEO performance indicator for an e-commerce site:
Duh. ‘b. Sales from organic search.’ gets you an A+ on this one. All businesses do much better when they’re actually making money. Listen to Mitt Romney &emdash; he repeats that every 30 seconds or so.
10: Pick the most important SEO performance indicator for a lead generation site
I didn’t even change the letter of the correct answer. It’s ‘b. Leads from organic search.’ Same letter as #9.
11: What are the first 3 things you check when doing an SEO audit?
Annoying, I know. There are no right answers to this one. On the bright side, there aren’t many wrong answers, either, unless you start writing stuff like “The keywords meta tag,” in which case I might slap you. Things like ‘duplicate content’, ‘broken links’, ‘canonicalization’, ‘title tag use’ and ‘graceful degradation’ are all good answers, though.
12: List five things that impact a site’s authority.
I wasn’t sure I could list 5 things when I started, but as it turns out, I can: Linking domains, quality of linking domains, relevance of linking domains, social media citation, canonicalization, redirect usage, link velocity, citation velocity…
You get the idea.
13: You got something totally wrong. What do you do?
NO FAIR IAN. You’re asking ethics questions now!!!
Tough shitskies, as my cycling coach used to say.
The best answer here was ‘b. Find the solution, then tell your manager what happened and how you’re fixing it.’ Ideally, it’d be up to you and the manager to talk to the client at that point. But if someone went with ‘c,’ I wouldn’t be horrified. ‘d’ is only OK if you’re me.
14: A client’s confused as heck after 3 emails. What do you do?
Please, for the love of all that’s holy, tell me you picked ‘a. Call them on the phone’.
15: What’s a fast way to add links…
… to a web site’s link profile that won’t violate the Google and Bing Terms of Service?
Answer a, ‘Buy links’, violates the hell out of both engines’ terms of service. Answer b is good, but a content campaign takes time. Answer c, ‘Fix broken incoming links’ is your best bet.
16: Define canonicalization
If you reference saints and Christianity, you blew it. See my article on canonicalization on Search Engine Land.
17: Describe blended search
It has nothing to do with a Cuisinart. ‘Blended’ search describes mixing images, videos, product results, local search and whatever else Google or Bing think might keep you around another .05 seconds into the traditional organic results.
I wrote a whole series of articles about this back when it was called Universal Search.
18: A client is panicked…
…They’ve just fallen out of the rankings for the phrase “foo bar”. That was their top traffic generator. They want to shut down all onsite content and technical SEO and focus 100% on links. What do you tell them?
No perfect answer here. But “HAHAHAHAHAH YOU ARE SO SCREWED” is not acceptable. Something involving sympathy and a game plan would really be better.
19: Tell me what you think PageRank is
This is more about me seeing how my team thinks than getting the ‘right’ description of PageRank. Although, if someone said “That little number in my Google Toolbar” I might get a little worried.
20: What’s a ‘headless browser’?
Man, I got some funny answers (many unprintable) for this. You guys are seriously twisted.
A ‘headless browser’ is, as Michael King put it, a web browser with no user interface. He wrote a great article about the whole topic here.
It’s relevant to SEO because there’s strong evidence Google uses a headless browser to render page layouts, thereby figuring out when you’re being all sneaky, using jQuery to hide all that ugly content in tabs and stuff.
a reverse an inverted index?
Also called an inverted index, a reverse index lets you efficiently store a keyword index of a document, by storing each word or phrase once, and then storing the numeric locations of that phrase. It’s cool. Really. I’m serious.
Although, once again, I screwed up. Technically a reverse index is another optimization technique, where you actually reverse the key for a piece of data. I won’t even try to explain that, because it makes my frontal lobe hurt. In future versions I’ll just make this question “What is an inverted index?”
I truly do suck at this. I could write standardized tests for school systems.
22: To you, the One Ring is:
Well, duh. The answer is A. Sauron’s jewelry. Read The Lord of the Rings if you didn’t know that, K? All Portent employees must know The Lord of the Rings, Doctor Who trivia or at least a bit about Star Trek.
SEO testing and training: Coming soon?
I have to say, this whole exercise makes me think there might be a way to at least teach SEO, if not test for SEO skills, in a relatively consistent manner. I’m noodling that a bit now. I’ll keep you posted.
You have given balanced answers to every question. Quite impressive!
Nope, don’t do it. But to be honest, you’re running an SEO company, and you should know that.
1. Sticking nofollow on links from user generated content can have a significant effect on how much Google trusts your website.
That being said…
Say your site has tons of nofollow links from user generated content
… and you then you somehow encrypted the whole link element server side into something like this
But you saw no benefit in overall search/authority (and no negative)
Where would you determine the pagerank that supposedly evaporates from nofollow actually goes?
Not sure I’m understanding the question, but when the PageRank evaporates, it really evaporates (as far as we know). It’s just gone.
When we’ve tried schemes like you’re describing – encrypting the link and then removing nofollow – it seemed to have a positive effect on rankings, which implies the PageRank was restored. These were only on test sites, though, so I can’t be 100% sure.
I wonder how much more I could qualify that answer?
Re: 2: When are drop down menus bad for SEO?
I’d be just as concerned with poor usability especialty on mobile devices. What is bad for the user will eventually be bad for SEO if it isn’t yet.
Yeah I’m not a fan of drop downs. I do, though, think there’s a right and wrong way to do them. Making them the dumping ground for everyone’s ego links is not the right way.
Entertaining as always. I’d like to see more kittens though.
Awesome – a few questions I didn’t actually know the answer to! Inverted index? Never heard about that one before … or at least when I was listening.
Why are you trying to attract pandas? Shouldn’t you be repelling them?
I really enjoyed the wrong answer choices. Nicely done.
I’m glad you decided to publish the answers. There were a few that got me – most notably the inverted index. Looks like you haven’t written about that (at least on this blog). Any chance you’ll have a post on that in the near future?
I might try – inverted indexes are complex enough that I can just understand ’em. Explaining ’em is a whole other matter.
“So, you’re tellin’ me I have a chance!”
-Dumb and Dumber
I hope you do write it at some point, lest I have to conduct the indecency of research on my own accord. A sad day, indeed.
Great SEO exam Ian. I wanted to elaborate a bit on the first question. I’ve noticed with a few test sites that a link profile with absolutely no “no-follow” links, comes off as unnatural to Google. It’ll start to rank well at first, but quickly seems to either get slapped, or discredit the authority pointing to the page.
I like to make sure that when I launch a new site, I start building a healthy number of nofollows. I’ve also found that “writing” the URL seems to help as well. Sometimes, when the platform I’m posting on doesn’t allow links, I’ll just write out the URL, http and all. I haven’t done and serious AB testing on this, so maybe I should shut my mouth, but the angle I take is to make the link profile as far from synthetic as possible.