Our imaginary link-building department is bustling with business: the Link Strategist created a campaign, the Content Designer bridged the gaps between brand and opportunity, and if the Prospector did her job right, most outreach targets should be right on topic.
But when you’re working at enterprise scale, “mostly on topic” isn’t close enough. So now it’s the Qualifier’s turn.
The qualifier answers the question, “Do we email the person behind this opportunity or not?”
The answer to this question is as fuzzy as your favorite kitten. It’s full of complications and considerations, which is why the Qualifier needs to be well-informed and well-trained and must come equipped with a healthy dose of common sense.
- Brand Protection. Qualification happens to protect your brand from looking foolish. As I’ll explain below, machines can fairly easily tell you which sites you shouldn’t reach out to, but only humans can verify which blogs or online linkers actually make sense for the brand. Reaching out to someone who’s a competitor, who’s already in touch with another department or whose interests are clearly off-topic with your content can create an awkward situation at best.
- To Save (Everyone’s) Time. Being respectful of our own time, as well as that of others, means that we’re not going to waste time and inbox space on pointless outreach.
How To Evaluate (Qualify) Pages At The Enterprise Level
In this article, assume you’re qualifying around 1,000 URLs.
Note: Metrics are not the sole way of qualifying opportunities.
Metrics can work against you. It’s like dating: if you don’t talk to someone who might be “out of your league,” or if you disregard people who seem “beneath” you, you may miss out. We’ve acquired excellent .gov links from small-town websites with a Domain Authority of 10 (out of 100); and we’ve placed guest posts on sites with a Domain Authority of 97.
Our model is based on relevancy rather than high Domain Authority, PageRank and/or similar metrics. Topical relevancy is a big deal. We pursue all the links where it topically makes sense.
A couple of years ago, Russ Jones wrote a post for Moz on what an organic backlink profile looks like, based on researching 500 random Wikipedia.org pages. The idea was that Wikipedia has many links and no link-building team — thus, their backlink profile is essentially 100% “natural.”
And it’s messy. Sites like Wikipedia don’t magically receive natural links from only sites with a Domain Authority of 50+, and neither should your site.
In this post, I’ll outline the processes for human and machine qualification and how to use both (along with a stellar Qualification toolbelt).
Human Vs. Machine Qualification
no’s can tell you what a page is linking to, but a human can tell you what a website curator cares about.
People often debate whether humans are machines are superior at a given task. For qualifying link-building prospects, however, I recommend using a combination of both. Humans and machines each have different strengths, and combining them is optimal.
What Humans Are Good At
- Data gathering, with certainty. Humans are good at specific types of nuanced data gathering that machines don’t “get.” With tasks like contact finding or website categorization, machines can make good guesses, but they’re going to be wrong sometimes. Thus, if you only use machines for qualifying, you can’t be certain that you’re reaching out to someone who will find your content or email valuable.
- Determining website and topic validity. Some websites are created to trick machines. URL structures or copy and design techniques that would immediately alert a human that something fishy is up can fool a computer into handing over the “legitimate” stamp of approval.
- “Getting” your brand. Again, it’s about nuance. Humans can be trained to understand a brand well enough that they won’t make awkward outreach mistakes. Machines can’t develop a gut sense needed for optimal qualification.
But I’m not advocating a Luddite link-building culture. We use a large suite of commercial and in-house tools at my company, many of which are used to help with qualifying.
What Machines Are Good At
Machines are better at doing a whole lot of stuff really freaking quickly. They can disqualify obvious “no’s” and take a whole lot of domains from “???” to “maybe” in the time it takes a Qualifier to refill his coffee mug.
It’s just that humans have to move opportunities from a “maybe” to a “yes.”
Opportunity Qualification By Hand
Note: Qualification needs to be campaign specific.
For example, if you’re trying to acquire links from local .org websites, but your client only wants links from sites with a minimum Domain Authority of 35 or 40, there may only be 10 websites that “qualify” for you. In this case, you may need to adjust your qualifier standards, either by topic or Domain Authority.
However, there are some universal questions that a Qualifier should answer for any campaign.
1. Is this a real website?
This is a human judgment made by exploring the website; is it run by people, or is it a scraper/machine-created website?
2. Is this a site we could reasonably get a link on?
Here, the Qualifier should know to eliminate sites that will require some exceptional circumstances (not created by a link builder) in order to obtain a link. These will vary for each campaign.
For example, it’s virtually impossible to get a link on CNN.com if you’re doing links page outreach, but if you’re doing PR outreach, the individual reporter at CNN.com who’s interested in your topic may be a great outreach target.
Quick exception: We recently got a mention on a government website. Honestly, if I’d been qualifying the list, I may have taken it out. But someone else made a different judgment call, and we got the link. So qualify, but don’t underestimate your content.
3. Is the page or domain on topic with our asset?
This is where you’re looking at that page and asking, “Does our content fit? Does it support the goals of the page?”
This is a page-by-page effort to figure out the topic of the page and determine whether or not pitching our asset makes sense for this page. (We don’t want to pitch our “guide to home buying” to a health linker.)
Example: We have a list right now of Spanish resource sections of .gov and library websites. The problem is that we’ve got individual domains with multiple Spanish language URLs. So which specific URLs should we pitch to? We know our Qualifier has to categorize them, based on related outbound links and the topics of each page.
Prospecting-Method Specific Qualifiers:
There are more than two ways to do prospecting, but there are two main ways we do it at my company, so that’s what I can speak to. For these, the Qualifier will need to work closely with the Prospector to stay aware of which methods she used to gather opportunities.
Co-citation is the prospecting tactic of finding two to four highly relevant and quality pages related to our resource and searching for sites that are already linking to at least one, maybe more of them.
Watch out for scraper sites and DMOZ clones.
The link graph tools out there (Ahrefs, Majestic, Moz) aren’t necessarily equipped to remove instances of duplicate content. Duplicate content on different domains is a major problem. For example, you might identify five different domains that are linking to the same page — but once you look at the pages themselves, you’ll see that the content and URL path are exactly the same in each case.
And keep an eye on freshness.
The web is a moving target. How recently has your backlink graph tool checked the page?
2. SERP Sourcing
SERP Sourcing is a prospecting tactic of finding resource pages and online linkers right from specific searches.
Watch for false positives and way off-base topics.
For example, if you’re looking for links pages, one thing you can do is look in the title of a page and scan for the word “links.” But “links” is a term that’s also used in golf. Or there can be news stories of things that are “linked” to a certain medical condition. And there are a ton of ways content confusion can happen when prospecting by machine. After all, cancer is also a horoscope sign.
Plus, depending on how thoroughly the Prospector searched, the Qualifier can end up with a lot of garbage. Payday loan sites can have crazy amounts of links pages because they’ve scraped someone else’s page. A given URL may be “on topic,” but do you really want to reach out to them? Is the domain something you want to be associated with? It’s worth going to the home page just to check out the site.
A human Qualifier knows these wrong opportunities when he or she sees them, but it’s necessary to take the time to look.
Check for outbound links on-page, as well; remember, not all linking pages explicitly say they are linking pages.
Final note on working with non-native speakers:
If you’re working with Qualifiers who are not from your culture, be prepared to explain cultural nuances.
Qualifiers need to know the brand and opportunity well. They need to know topically why an opportunity is valuable to its recipients. People who don’t know the big picture (“This would not work for content A, but it might work for content B.”) can’t qualify with outreach certainty.
Opportunity Qualification By Machine (In Bulk)
You can’t ask a machine to know what is an opportunity, but you can dictate what isn’t. For example: “Adult” stop words.
At Citation Labs (my company), we have some disqualifiers for blogs because we have times when we don’t want any blogs on our list (see the Stop Words in our toolbelt for examples). It comes down to tactic and content.
For machine qualification, I would recommend getting familiar with regular expressions. Use regex for removals in a spreadsheet — or, if you want to get fancy, getting someone to build you a tool. (Or just wait a few months. We’re currently working on a suite that will enable this.)
For machines, just as with human qualification, what is and isn’t a link opportunity can vary by campaign. From a pure machine perspective, that’s why there aren’t any hard and fast rules. We prospect so thoroughly sometimes that we get some spam pages, so if the word “nike” appears on a page, it’s a definite “no,” unless we’re doing outreach for a footwear campaign.
For each campaign, it’s important to know your definite NOs:
- Certain TLDs, primarily non-country TLDs. Except for when they make sense. When you’re allowing in pages that most SEOs don’t (because of country or Domain Authority), you’re going to be in a less crowded inbox.
- Certain string occurrences. For example, “alexa” or “wiki.” We call these “stop words.”
- Who’s on the “Do Not Contact” List? Make note of your competitors or websites that are currently in contact with other departments. Always ask the PR department. They’re going to be really touchy about who you reach out to.
As Promised: The Qualifier’s Toolbelt
Even human qualification can use some machine help. Below are some of the tools we find especially useful at Citation Labs, along with a sample list of “stop words” we use for some machine-based qualification campaigns.
- The “super deduper” for deduplicating
- The List Comparison Tool
- Sample Stop Words:
Qualifying isn’t always sexy, but it’s an essential safety net for every campaign. Without it, we’re ruining brand trust left and right and leaving the Relationship Builder to clean up the mess.
Speaking of which, our Relationship Builder is caffeinated up and waiting in the wings. He’ll be the star of our next and final post in this series on the Enterprise Link Building Team.