Building Links with Multiple-Citation Analysis
By: Ben Wills |

We're getting pretty excited here at Ontolo. We're putting the finishing touches on V3 and are looking forward to rolling it out on Tuesday, June 28th.

The past couple of days, we've written about how to find more valuable links with your link building campaigns. Hearing peoples' comments and feedback, it sounds like a lot of you are starting to shift how you think about your link prospecting strategies. You're beginning to see how it's possible to build more valuable and more relevant links, much more quickly with the new Toolset that's being released next week.

Today, we're taking it another step further.

We're discussing a topic that's thrown around casually and in ways that, most of the time I hear reference to this concept, it's obvious that it's misunderstood and under-utilized. A large reason for that is that it's a “neat idea” and that the tools simply don't exist to execute this kind of strategy well.

The concept is: Multiple-Citation Analysis.

The fundamental purpose of Multiple-Citation Research in link building is to identify pages or sites that link to at least two other websites that you define. Typically, in link building, those two (or more) websites are your competitors. The idea is that if there is a web page or a web site that has linked to two competitors or more, that they are much more likely to link to you.

It's a fairly straightforward concept. But we see almost no one using Multiple-Citation Analyses effectively.


The best way we can demonstrate why no one is performing this well is to walk through specific data on a project we ran using pure Multiple-Citation Analysis with the tools that currently exist in the marketplace: SEOmoz and Majestic SEO. (Later, I will show you howyou can quickly and easily perform highly-effective Multiple-Citation Analyses.)

Details of the Multiple-Citation Analysis Project

  • 4,000,000+ Citations were analyzed.
  • Backlinks for 15 competitors were included.
  • Citations were analyzed in contexts of URLs, Subdomains and Domains.
  • Citations were counted only if they met minimum quality scores as defined by Majestic SEO/SEOmoz.

Our analysis only included sites with a certain minimum trust score in SEOmoz or Majestic SEO. We put every URL and citation into a database, then analyzed where there were links to competitors from a URL, subdomain or domain.

A Word of Caution

If you undertake a Multiple-Citation Analysis project like this on this sort of scale, be very well prepared. This project requires efficient code, a well-optimized database, and the hardware necessary for this kind of analysis to take minutes instead of days. Without the right preparation, resources and code, the time to import and analyze the data may often be more than 24 hours of run time. Testing and debugging this code was quite interesting to say the least.

But. Would You Like to Have the Code We Wrote to Perform Your Own Multiple-Citation Analyses?

We're considering releasing this code as an open-source project. The reason why we haven't considered releasing any code as open source before is because it takes a fair amount of time and energy to setup, promote and maintain an open source project. That said, we're open to it if it's something the SEO and Link Building Community would really find useful to have.

If, by the end of today, 100 people or more click the link below and Tweet the enclosed message, we've decided that we will release the project next week a couple of days after we release Version 3 of The Link Building Toolset.

So, if you are interested in it or if you believe the SEO and Link Building Community would benefit from it:
     - Tweet a Vote to Release the Multiple-Citation Analysis Code as Open Source

And now, back to the data.

Results of the 15 Competitor, 4,000,000 Citation Multiple-Citation Analysis

Citations URLs Subdomains Domains
2 20,619 12,755 10,818
3 2,483 6,197 5,624
4 381 2,636 2,554
5 101 1,149 1,131
6 46 498 159
7 17 262 481
8 10 148 262
9 0 76 83
10 0 47 48
11 0 29 36
12 0 8 11
13 0 1 3

20,000+ Prospects Looks Like a Pretty Solid List, Right? Or is it?

Let's take another look at the data, and put the numbers into percentages. I've taken each total, divided it by 4,000,000, then presented the total citations found as a percentage.

What each number in the chart below represents is the percentage of ALL citations analyzed which have that many citations present on a URL, Subdomain and Domain basis.

Citations URLs Subdomains Domains
2 0.515475% 0.318875% 0.270450%
3 0.062075% 0.154925% 0.140600%
4 0.009525% 0.065900% 0.063850%
5 0.002525% 0.028725% 0.028275%
6 0.001150% 0.012450% 0.003975%
7 0.000425% 0.006550% 0.012025%
8 0.000250% 0.003700% 0.006550%
9 0.000000% 0.001900% 0.002075%
10 0.000000% 0.001175% 0.001200%
11 0.000000% 0.000725% 0.000900%
12 0.000000% 0.000200% 0.000275%
13 0.000000% 0.000025% 0.000075%

So what does the data above tell us?

A Multiple-Citation Analysis using 15 competitors and over 4,000,000 citations demonstrates that the most basic, fundamental, lowest-quality occurrences of Multiple-Citations are in approximately 1 out of every 200 competitor backlinks.

With the current tools that exist today, Multiple-Citation Analysis is a relatively low-performing method of link building.

But. There Is A Better Way.

The problem that we found over and over with multiple-citation analysis projects is that, even though it might appear that because a URL or site links to multiple competitors that it's a good link prospect, the reality is that it's simply not.


Because there are a lot of sites out there that "categorize" domains and link to various sites who are "similar." In your analysis, you have to weed through all of those. Here's an example of one that's quite popular, showing the data for SEOmoz: SEOmoz on AboutUs.

Another reason? Automated directory sites. It's a similar problem to the example above.

Between those two, that 0.5% of multiple citations being considered as link prospects significantly decreases.

So why bother?

The opportunity for effective Multiple-Citation Analyses lies in combining your strategy and analysis with relevance factors as well as value and quality factors.

Incorporating Relevance into Multiple-Citation Analyses

What we've done with the next version of the Ontolo Link Building Toolset is put the power in your hands to fully define how your co-citation and multiple citation analyses happen.

Here's the step-by-step for combining Multiple-Citation Analyses with relevance searching:

1. Define and Input Your Competitors

* Hint: Use the SERP Dominator to find your top SEO competitors.

2. Choose Which Competitor Citations to Include

3. Define Your Minimum Citation Requirements

ie: If you were to input "2," then you would only get results back that linked to two or more of the competitors you have selected above. 3 returns three or more, etc.

4. Define Your Search Query

And / Or
And / Or

Relevance Searching + Multiple-Citation Analysis: Leveraging The Results

What the settings above do for your Multiple-Citation Analyses is significantly improve your results to the degree that the results actually become useful. Without the level of relevance as outlined above, pure Multiple-Citation Analysis still leaves you in the dark. Even the results that contain 10 or 11 citations to competitors...we were hopeful that they would yield positive results, but each one simply resulted in nothing useful.

When we did finally take all of the URLs with multiple citations and put them into a content-searchable database, we found the vast majority of them to not be useful. In fact, if we hadn't applied any keyword search terms to our prospecting within that list, we would have hit a similar 1 out of 20 ratio that I discussed yesterday. It was still simply unacceptable.

BUT, the best way to use multiple citations in your link qualification is to have a specific kind of link or targeted link content that you're looking for.

An example of this is "top 10 lists." If you have an iPhone game and you're looking for top iPhone game lists, use sites like Angry Birds, Hanging with Friends, Words with Friends, and Fruit Ninja (all top games right now) for your Multiple-Citation Analysis AND search your prospects for "iphone games." That kind of list - prospects that link to other iPhone games AND use the phrase "iphone games" - is going to be a much higher quality list that gets you many more links in a lot less time.

And, it brings you right back around to the purpose of Multiple-Citation Analyses anyway: If they've linked to your competitors, they're more likely to link to you.

Multiple-Citation Analysis allows you to leverage the principle of Commitment and Consistency in your link building campaigns, resulting in significantly higher link acquisition rates.

In short: More Links, Less Time.

Post a New Comment