Links Matter: How to Measure and Attain Them – BlueGlass LA
Hey there! It’s after lunch on day 2 of BlueGlass LA. Mysteriously, all the people who were missing this morning have finally shown up. Was it because they heard who the speakers were? I’ll bet. You’d shake off your hangover too for Rand Fishkin, CEO and CO-Founder, SEOmoz and Dave Snyder, Partner, Search & Social (we hear they’re called BlueGlass now but I guess the conference website hasn’t caught up. Whoever the parent company is should really get on that.)
Also it’s freezing in here and there are record numbers of people huddling over a cup of tea like its their Precious. I think Winfield is trying to kill us.
Rand is up first. His 800 slide presentation is called Strategic Link Analysis for SEO.
Step one: Determine your goals – What do you want to achieve with links? Certain kinds of links will move the needle in different ways.
Types of link building goals
- Individual, competitive rankings
- Greater indexations
- Improve a site’s overall ability to rank pages
- Dominate the SERPs – controlling SERPs in multiple verticals in order to cover page 1.
Step 2: Match Goals to Competitive Analysis
Individual Rankings: What do you need to rank here? How hard will it be to rank on page one. Look at how popular and important are the top ten people ranking are. Look at Domain Authority, # of domains linking Root domain, Page authority, # of domains linking to page, # of partial match anchor links, # of exact match links.
Step 3: Identify the Missing Pieces and Create a Plan
Indexation (example) How do I get Google indexing more of my pages?
ID sites with high indexation and compare metrics (particularly in the same industry.) how hard is it?: # of pages indexed divided # of linking root domains
Domain Rank-ability: How do I get to rank like Wikipedia ranks? Execute a link acquisition campaign centered on missing metrics.
SERPs Domination: Run analysis of current SERP as well as where your current pages rank. Also do a vertical analysis. Videos are being heavily promoted in Universal in the last 6-12 months. News, Blogs, Updates = yes. Books = maybe. Images?
Create Content/Profiles on Powerful Sites. Target link sources with flexible applications. Take advantage of bios.
Figure out what metrics you care about. If it’s raw link juice (and that doesn’t mean PageRank alone.) Link source quality, link quantity + domain diversity, anchor text.
Save yourself a ton of work by research & building a plan before the first link is targeted.
And now, after a flashy false alarm from the fire alarm, it’s time for Dave Snyder and linky goodness.
The key to dominating the Web is links. They’re the core of the algorithm because they’re the core of the Web. It comes down to traffic. Links carry more value than just their ability to help you rank. They are the way people found and shared content even before the invention of search engines.
PageRank made them into a commodity.
So if you want to get traffic or income on the Web, you have to get links. Link acquistion boils down to two concepts:
1. Monetary Response — Don’t YOU do it. Leave it to….really good SEOs. [sigh — I’m really just reporting here, folks.] This is simple, people love money and will do disgusting things for it. However, you’ll get caught
2. Emotional Response — Robert Plutchik identified eight primary emotions: anger, fear, sadness, disgust, surprise, anticipation, acceptance, joy
People link to things that they respond to emotionally: Wikipedia = trust, conspiracies = fear, BP = disgust, etc.
Initial point of contact > Thoughts > Feels > Actions > Result
Example: Vegan support group poster > carnivore sees it > reacts > goes back to dorm and makes a parody > posts parody by original poster
You don’t need to start from the initial point of contact. You need to work back from the action. Why would people link to me, where do I want to get it from? what feeling or emotions will support those actions, how do I trigger those thoughts that trigger the feeling? what content will trigger those thoughts.
Market research is important. You must understand the audience you’re creating content for.
High level link process:
Step 1: Mine Data
- What are people linking to in your vertical? (Linkscape)
- How are they linking? (Linkscape)
- What topics are people emotional about now? (Google/Twitter Trends)
- How have people historically interacted with content?
Step 2: Create Your Reason for Linking
- What emotion do you want to evoke?
- How can you shape this info into content that not only get traction but also will evoke the emotion you’re looking for?
Step 3: Craft outreach strategy from data
- create an outreach list
- look at social venues that have created links in the past
- make lasting relationships within your vertical that can be quickly and easily leveraged based on trends
Step 4: Outreach
- position your outreach to match the emotional response you intend to receive
- post-QA on all acquisition (check out his article on quality factors on SEJ)
- quantify and inventory the links
Step 6: Data collection and categorization
- look at outreach metrics like opne rates on emails, acceptance rates
- QA links that come back in
- Segment links based on value and percieved value in the space
- What worked, what didn’t?
Step 7: Rinse and repeat
Google is smarter now. They used to take out sections of the link graph, now they’ll target individual links. Focus on “will this link increase my traffic and share my content with the correct audience?” Google has a camera in your house Right Now. They’re using Analytics, Toolbar, etc. Make the links valuable for things other than SEO.
We have a large, deep-content site, many of which are not so high quality. Something something pagerank sculpting?
Rand: Large sites like yours see the most value from PR sculpting. But in your case, can you scale and extend the good content? I’d just ignore the bad stuff for now. See if you can just get a couple of links to deep level pages.
Dave: Traffic can help how fast and how deep a site gets crawled. Build a following. The social signals part is so underestimated right now.
Have you found any difference in ranking TLDs?
Rand: Yes but not for the reason that we might associate. Google doesn’t have a bias against .info or .net. I think that exists because of human bias more than search engine bias.
Dave: Stick to .org, .net, .com just from a conversion standpoint. I’ll never buy from .biz, they’ll probably steal my credit card. Every decision about your site can’t just be made based on one factor. It can’t just be SEO.
How do you diagnose a drop in rankings?
Dave: If you didn’t change anything, go look at your link profile and see if that changed, if anything got gray barred or disappeared or those pages changed. If it’s not you and it’s not the links, it’s the ranking systems.
Rand: Most of the time, it’s because you had links that were wiped about because they weren’t quality and they were taken out by a ranking change. Focus on getting good links. One link can make all the difference.
How important is exact match anchor text?
Dave: We take a holistic approach. Don’t worry about hitting every single metric every time. You can live without some things if there are others.
Rand: If the links you are acquiring not just natural but the kind that Google will always want to count, go for the exact match anchor text. If someone’s picking up a badge, that’s natural. However, it seems like today exact match anchor text is weirdly overpowered.
Dave thinks it’s more how you get the link than anything. Google will kill something you paid for and exact match can be a signal there.
Do aged links count more?
Rand: I suspect, my personal opinion is, no it doesn’t matter much. Fresh links seem to provide a spike in value and then averages out. Pure speculation. [YMMV, void where prohibited, etc]
Dave: I think trust plays a role in ranking, so site age does seems to matter. But don’t go only build links from old sites if you’re brand new. It shouldn’t just be “this is how we build links”. You need to take a holistic approach.
Do tweets get you to rank as well as indexed?
Rand: When they looked, they did not see it influence ranking BUT it was two weeks after their experiment that the tweets started to get indexed. So, he thinks, yes, that it probably affects it some.
Dave says he thinks it does short term, because of quality deserves freshness. Jump on trends and try to figure out how you can tie into it.