Google-Safe External Linking

One of the consequences of Google’s Panda and Penguin updates is that people are afraid to link out to third-party sites. Nobody wants to inadvertently point to a site that is engaging in practices that violate Google’s terms, which in turn might result in a penalty to the linker. Unfortunately, the guidelines from Amit Singhal aren’t sufficient as an evaluation tool, as is proven every time a big brand (a site the general public would “trust”) gets hit with a penalty.

However, linking to a site and indicating your trust of a site with that link is part of being a good web citizen. And so the knee-jerk reaction of “nofollowing” all external links isn’t ideal, but what else is a site owner supposed to do? Evaluating an entire site and the activity of its owner(s) just so you can feel safe about linking to a page is a ridiculous requirement that no one has time for. And even if such efforts were undertaken, the evaluation would be accurate for only a short period of time since the very next day the site’s owners could decide to be overly aggressive with their SEO efforts.

Method 1: Third-Party Traffic Data

It’s reasonable to expect that a site hit by a Google penalty will see lower traffic levels in the following weeks. And often this decline can be seen by looking at data from My own testing has shown that penalties do show up in Compete’s free charts, but unfortunately there are a lot of false positives due to the free charts including all sources of traffic; not including enough data to adjust for seasonality; and there being no obvious way to identify special promotions (like a sweepstakes) that have come to an end.

Method 2: Third-Party XYZ Rank Data

SEOmoz (now just Moz) is probably the best known provider of a PageRank-like score. It provides three domain-level metrics: Domain Authority, Domain MozRank, and Domain MozTrust. For relative comparisons between sites, these metrics can be informative. The challenge with them is that no matter how strong the math or how comprehensive the index, these metrics aren’t an exact match for Google’s calculations, especially when it comes to a site that has been hit with a penalty.

So maybe you’re inclined to turn to Google’s Toolbar PageRank as a way to identify a domain that is safe to link to. The shortcomings of this approach have been covered repeatedly, so I won’t rehash them except to say that Toolbar PageRank isn’t necessarily up to date and doesn’t necessarily reflect any penalties.

Method 3: Using Keyword Rankings

This brings us to an unlikely option: keyword rankings. After all of the proclamations that ranking data isn’t important and that ranking reports are dead, who would’ve guessed that the oldest method of reporting would be used to detect the impact of the newest algorithm changes? Enter SEMrush.

Using SEMrush works (mostly) for two reasons:

  1. Keyword rankings aren’t seasonal, so month-over-month comparisons are fair.
  2. The data is isolated to Google’s organic listings, so activities in other channels won’t have any effect.

Caveats: I’ve seen SEMrush’s data not align with what’s reported by Searchmetrics or Sistrix, so be careful with your conclusions. However, for sites where I absolutely know there’s been a penalty, the graph I’m about to highlight has also shown declines. Or to put it another way, I haven’t seen any false positives, but there have been false negatives.

Here are the steps to get to the graph I like to look at:

  1. Type in a domain and click on the Search button.
  2. Click on the SE Traffic heading of the top-right chart.
  3. Under the domain there’ll be a drop-down that you should set to Total Keywords. Click on Plot Chart to refresh the data.
  4. Expand the date range so that it starts on Jan 2012 (an arbitrary pick really, but it gives you a big-picture view). Click on Go to refresh the chart.
  5. Compare the sudden dips with the schedule of known updates.

Figure 1: SEMrush reveals a couple of declines potentially caused by Google penalties.

I’d say that there have been at least two penalties to the above site. One in November 2012 (thought by many to be a significant update) and another in May 2013 (Penguin 2.0).

If you were originally considering linking to the above domain, would you still do so?

Related reading

Google home page