.edu Link Authority Introduction
A couple weeks back, one of my colleagues at SeatGeek, Max Fram-Schwartz, was featured in Macalester’s Classroom to Conference Room’ series. This kicked off a discussion in which we were wondering what would be the ideal .edu domain in which to get an inbound link to your site (holding other factors equal), from an SEO perspective (i.e. .edu links domain authority). We began to scurry off to our respective college websites and check their domain authority via SEOmoz.org (if you are new to domain authority you can think of it as PageRank, but across an entire domain). The few we checked were fairly close, but still far enough apart to be significant given the logarithmic nature of authority, so my curiosity was immediately spiked and I decided I was going to turn this discussion into a blog post (charts below).
Approach / Thought Process
- PageRank’s origins are partially derived from the idea of citations in academic papers
- Papers receiving the most text citations are essentially the most authoritative on a given topic
- .edu domains are not innately high in authority, but rather tend to be authoritative sources based on how well they generate quality inbound links, produce unique high-level content, etc.
- I started operating under the assumption that academic studies from top universities such as MIT and Harvard are cited the most often in offline papers
- From this original assumption, I concluded that the top universities should also be the most authoritative online in terms of SEO domain authority (better university, more links/online citations, more authority…)
- I looked at the top 50 schools according to US News Best Colleges 2010 – National Universities
- I documented the SEOmoz domain authority of each, along with the number of inbound links they have generated and linking unique root domains
- I ran some basic analysis on the rankings and the output is below
SEO and .edu Domains: The SEO Rankings
To put the scale of authority in perspective, only the web’s top domains like google.com and yahoo.com are above 90, amazon.com is in the 80s, market leading e-commerce sites tend to be in the 60s or 70s and many well established web-based companies are in the 50s
Notes on the SEO rankings:
- MIT.edu is the clear stand-out at the top, with a domain authority of 78 and 38.1 million inbound links
- As a reference point, I looked at a second tier school, Drexel, (solely because that’s where Chiddy Bang went) and Drexel’s domain authority is 56
- USC.edu is staggeringly low and the clear loser among these top 50 universities (domain authority of 50)
- USC has a solid number of inbound links from unique domains, so at this point it’s low ranking is a mystery (I will possibly write a future post looking into this and whether or not they are committing serious SEO mistakes)
- Rounding out the top 5 in domain authority are Texas, UPenn, Carnegie Mellon and Berkeley
I quickly looked into one of the key drivers, number of inbound links, and have charted the 5 colleges with the most inbound links below (note the median number of inbound links is 5.8 million among the top 50 schools). For those that don’t know, inbound links, in their most basic form, function like votes for a website’s credibility/relevance, so a higher number of links (all else equal) will yield better domain authority.
Make note of NYU, Harvard and Washington, as their success here will come in to play later when we look at specific SERPs.
Identifying Outliers and Over-achievers Among .edu Domains
Once I had the Top 50 national universities compiled in a spreadsheet along with their corresponding domain authority ranks, I took the difference between their SEO rank and US News rank to generate a list of what schools were over-performing and under-performing the most from an SEO perspective (i.e. Texas’ SEO rank is 2, while their US News rank is 47, yielding a +45 over-achiever score and a top ranking in the chart). The results are below; a particular focus should be given to the outliers in the top 10 and bottom 10.
Notes on SEO .edu Domain Over/Under-achievers
- Texas ranked at the top with a +45 differential, while Johns Hopkins is at the bottom with -34 differential
- 8 of the top 10 are public schools, including 5 California public colleges
- There is trend where public schools perform strongly in terms of SEO rank and will call for follow-up analysis (particularly around school size and graduate program size)
- Despite the prominence of public universities, Carnegie Mellon (with a student body of in the 5000s) ranks 5th on this list and likely speaks to the strength of tech oriented schools on this metric (note MIT ranks so high for both, that it can’t have a high differential – definitely one flaw in looking at it this way)
University Domains Keyword Research
Once I had identified some interesting rankings and trends, I wanted to see how this surfaced itself in real search results. I did some basic keyword research to see the most searched for high-level terms related to colleges (note the brackets indicate these are exact match searches).
.
.edus in the SERPs
“university” search query notes:
- NYU (2nd in SERPs) is 3rd in number of inbound links and 5th in number of linking root domains
- Columbia (3rd in SERPs) is 1st in number of linking root domains (41,548)
- Harvard (4th in SERPs) is 4th in number of inbound links and 8th in overall domain authority
- Texas (5th in SERPs) is 2nd in overall domain authority, 2nd in linking root domains (34,732) and 1st in the SEO Rank v. US News Rank differential
- Washington (6th in SERPs) is 2nd in the SEO Rank v. US News Rank differential and 5th in number of inbound links
- For the search query,”college”, The City College of New York (6th in SERPs) and Harvard (7th in SERPs) were the only schools on page 1
.edu Links & SEO Conclusions
- Overall, top schools have really strong domain authority, and this difference really comes in to play when you look at second tier schools (i.e. Drexel at 56)
- Academically strong public schools perform really well in terms of SEO domain authority and ranking in SERPs
- Just because a school is large and public, does not mean it will have a high domain authority (Penn State scored a 53 and a sampling of several other public schools ranked in the mid 60s)
- These are all relative, and almost any link from a .edu domain is going to be solid as I have yet to see a school below a domain authority of 50 and most rank in the 60s or higher
For my next steps, I am going to look deeper at the impact of school size, graduate program size and more statistical correlations between the “goodness” of a school and it’s SEO authority. I will also have to take a look at liberal arts schools, I quickly looked at Williams (often considered to be the best liberal arts school) and they scored a 66.
I know their are probably several unsound assumptions or statements I made along the way here, so feel free to criticize or make suggestions in the comments section.
Known Issues:
- Assumption that top colleges receive the most citations in academic papers is not yet validated
- The decision to use US News Rankings is somewhat arbitrary and skews the output of the ‘Over/Under-Achievers’
- SEO data is only from one source, SEOmoz











Hey Chad – terrific analysis; I think this sort of thing is fascinating!
One good thing to be aware of – Domain Authority as a metric tries to predict how well any given page will perform in Google.com’s search results for a query. Thus, large sites that have many pages that don’t perform well will be dragged down vs. smaller sites that have more compact quantities of pages and thus fewer chances of ranking poorly with those pages.
We’re working on ways to remove some of this bias while simultaneously being highly predictive of rankings with DA, but in the meantime, you might also want to look at analysis of # of linking root domains, one of the metrics we like best (and, as a single metric, one of the highest correlated with good rankings).
Feel free to drop us a line, too!
Best of luck,
Rand
Thanks for the feedback, it is very much appreciated! I have all the raw data pulled for number of linking root domains and alluded to it in a couple cases. I am planning on doing a follow-up with that data. The reason I did not include all of it here was that I was seeing some anomalies in the data that I wanted to double check. For example, the median number of linking domains among the top 50 schools was 7,789, which made sense, however Harvard and some others were less than 400, which just seemed odd to me.
Thanks again. Maybe I can do the follow-up on the YOUmoz blog to get feedback from the whole community.
Thanks again, Chad
What about including an online university in the analysis? I can’t imagine University of Phoenix is cited frequently (if at all) by academic papers, but they must have quite a bit of unique content online compared to some traditional universities.
JH, I thought of that as well (plus anytime I get to do research involving Phoenix Online, I am happy) and their SEO domain authority score was 67.
I do not necessarily agree with you that they would have more unique content, and I would also question the quality.
Great read, Chad.
Have you seen Vistaprint’s new SEO tools?