Using a Clean Link List
It’s important to begin a link project by using a cleaned list. Use the following considerations to eliminate as much manual review as possible.
- Review only one link from each domain.
- Have a header checker test for links that have a server response code of 4xx, 5xx, or 302. These do not need to be checked in the priority batch. (You may want to check these later since sometimes changing the user agent will yield a different response)
- Test links with a tool like Majestic or AHrefs to identify the ones that use nofollow on them. These do not need to be checked.
The “Smell” Test
Take a look at the URL itself. If the domain displays any of the following, it will require a close manual review:
- It is a freehost like wordpress.com, blogger.com, or 000webhost.com
- If it is on a subdomain
- If the filename contains the words “link”, “submit”, or “resource” – these are indications your link is on a links page with a lot of other unrelated sites.
If the domain doesn’t contain any of the above, it still has to be reviewed, but domains that do contain the above should have a higher threshold of approval.
Check the Anchor Text
If the anchor text (a field available in the spreadsheet) is an exact match keyword like
- “payday loans”, “best red widgets” or similar;
- “what”, “this” or another unrelated word other than one that would indicate a natural link like “click here”;
- no anchor text (likely an image) or extremely long anchor text (like an entire paragraph);
- nonsense words or phrases;
… it should be given a higher threshold of approval. This does not mean that a link to your brand name is not ok; just that links that contain “exact match anchor text” or unrelated anchor text should be met with an extra dose of skepticism.
Do a Manual Check
- Open the URL.
- Use a plugin like SpamFlag for Chrome to see instantly if the link is nofollow and to find its location on the page. It looks like this:
The “Follow” 1 indicates there is one link on the page that is follow-able that points to the domain we requested.
The link above to CNN Newsource is the one we’re looking for. Clicking on the “1” above will shift the screen to this link so you can see it.
- If the link is nofollow, you can mark it as such and move on – it’s not one we need to focus on for this exercise.
- If the link is follow-able, move on to the next step.
- If you are not able to find the link on the page, flag it for follow up and move on.
Observe the Link Environment
Now comes the hard part. You have to make a judgment call as to whether this link is a valuable link or not. If any of the following are true, it’s probably not a valuable link. If the link falls into any of the categories with a * next to them, flag it specifically for further investigation.
- It appears in a comment or feedback section and is not part of the main page.
- It is embedded in a forum post or a forum signature.
- It is part of the author signature.*
- It’s included in an article that isn’t about a similar topic or doesn’t make sense.
- The article or post is written poorly, uses other random links, or is hard to read. Often may look like it’s written by someone who doesn’t have English as their first language.
- The link is contained in a footer, a blogroll, or in the main template of the site such that it appears on every single page.
- The link is completely unrelated, grouped with other unrelated links, or seems nonsensical on the site. Ex: link to an electronic cigarette site on a postpartum depression site.*
Observe the Site Itself
What type of site is it? If it’s any of the following, it’s probably a bad link:
- It only contains a few articles/posts
- It’s a site that allows you to “submit an article” or sign up to provide content (note, there are some sites that are ok for this like Wikipedia. Use your best judgment.*
- The article is about ecigs or a related topic to your site, but the other articles on the site are not related.
- The last post was more than a few months ago.
- The site seems like it was designed solely for the purpose of ranking for a keyword, and offers no other value.
- The site is a collection of bookmarks or associated with a specific userid.*
- It is a directory of pages where there are dozens of links on a single page. Note that sometimes these are legitimate, so use your best judgment. Generally, the directory is “bad” if it:
- References PR or PageRank
- Provides a link that is more specific than the website URL or company name
- Uses your DMOZ/ODP description
- Looks unreputable (i.e. 23dddedirectory.com)
Links for Further Investigation
Once you’ve done this sort of “triage” on the links that are in your “clean” list, go back and dig deeper on the ones you starred for further followup. These generally point to a pattern of bad links, such as article spinning, bookmarking spam, or paid link manipulation. If you don’t know what any of those things mean, hire a link expert. We’re expensive, but well worth it to uncover patterned practices like this that could hurt you far more and cost you far more money if not resolved.
Now that you have some idea of how to evaluate a link, feel free to download my link management spreadsheet (free) and get to work! You can hit me up on twitter at @jennyhalasz if you have questions.
(this post first appeared on jlh-marketing.com and is republished with permission.)