RNS SEO: How 52 research networking sites perform on Google, and what that tells us

Research networking systems (RNS) like Vivo, Profiles, SciVal, and Pure are meant to be used — but often fail to be discoverable by real users because of poor search engine optimization (SEO).

That’s why we’re releasing RNS SEO 2015, the first-ever report describing how RNS performs in terms of real-world discoverability on Google.

1. Methodology

We picked 52 different RNS that matched the following criteria:

  • associated with a single institution (excludes trade groups, collaborations, etc.)
  • based in a majority English-language locale (excludes France, Germany, etc.)
  • meant to be accessible to the public (excludes systems behind a firewall, not on port 80, etc.)

Here’s how we did it:

  1. For every single system, get a list of every single profiled user and the associated URL (e.g. “Kevin Labar” at https://scholars.duke.edu/display/per0978462)
  2. For every system, select 500 people at random (under 500 users in a system? select them all)
  3. For each name, search Google for “First Last Institution” (e.g. “Kevin Labar duke”)
  4. See if the RNS profile page shows up among the top 3 search results

We looked at search rankings for 24,583 profile pages across 52 different RNS sites, using the RankTank keyword rank checker tool to do bulk checking.

2. The Results

We’re measuring search engine optimization success by looking at what percent of a site’s profiles appear within the top 3 Google results for “First Last Institution”

  1. University of Melbourne 95% [VIVO] [under official domain]
  2. University of Massachusetts 91% [Profiles] [under official domain]
  3. University of California, San Francisco 87.8% [Profiles] [under official domain]
  4. Cornell 80.6% [VIVO] [under official domain]
  5. University of Colorado Denver 76% [Profiles] [under official domain]
  6. Boston University 73.8% [Profiles] [under official domain]
  7. Harvard University 71.8% [Profiles] [under official domain]
  8. Georgia Regents University 66.8% [Pure]
  9. University of Minnesota 64.2% [SciVal Experts] [under official domain]
  10. Penn State 62.6% [Profiles] [under official domain]
  11. Michigan State University 59.2% [SciVal Experts] [under official domain]
  12. Northwestern 55% [VIVO] [under official domain]
  13. Northwestern University 54.6% [SciVal Experts] [under official domain]
  14. Ohio State University 54.4% [Pure]
  15. University of South Africa 45% [SciVal Experts]
  16. Clinical Translational Science Institute at Children’s National 43.7% [SciVal Experts]
  17. University of California, San Diego 42.4% [Profiles] [under official domain]
  18. University of Colorado Boulder 42.2% [VIVO] [under official domain]
  19. Stanford University 38.4% [Custom] [under official domain]
  20. Wake Forest 38.4% [Profiles] [under official domain]
  21. University of Maryland-Baltimore 37% [Pure]
  22. University of Iowa 35% [Custom] [under official domain]
  23. University of Southern California 34.6% [Profiles]
  24. University of Illinois – Chicago 29.6% [SciVal Experts]
  25. University of Hawai‘i 26.9% [VIVO]
  26. Case Western Reserve University 23.4% [SciVal Experts]
  27. Thomas Jefferson University 23.4% [Profiles] [under official domain]
  28. Johns Hopkins University 22.2% [SciVal Experts]
  29. Scripps Research Institute 20.4% [VIVO] [under official domain]
  30. University of Nevada, Las Vegas 20% [VIVO]
  31. Wayne State University 19.2% [SciVal Experts]
  32. Washington State University 18.2% [SciVal Experts]
  33. University of Pennsylvania 17% [VIVO] [under official domain]
  34. University of Montana 15.6% [VIVO]
  35. Western Michigan University 15.4% [SciVal Experts]
  36. Duke University 15.2% [VIVO] [under official domain]
  37. University of Rochester 14.6% [Profiles] [under official domain]
  38. University of Florida 14.2% [VIVO] [under official domain]
  39. Montana State University 12.9% [VIVO]
  40. University of Nebraska 12.8% [SciVal Experts]
  41. University of Nevada, Reno 12.3% [VIVO]
  42. Temple University 11% [Pure]
  43. Northern Arizona University 10.8% [SciVal Experts]
  44. University of Arizona 10% [SciVal Experts]
  45. Arizona State University 9.2% [SciVal Experts]
  46. University of California, Davis 8.8% [SciVal Experts]
  47. University of Miami 7.4% [SciVal Experts]
  48. Albert Einstein College of Medicine 4% [SciVal Experts]
  49. University of Utah 2% [SciVal Experts]
  50. Texas A&M 1.2% [VIVO] [under official domain]
  51. Indiana University 0.6% [SciVal Experts]
  52. Oregon Health & Science University 0.6% [SciVal Experts]

The top sites have incredible search engine optimization — fourteen sites have scores of over 50%, and five with over 75%. On the flip side, twenty-three of the sites have scores of 20% or below, leaving lots of room for improvement.

3. Lessons Learned

Which software should you pick? It’s possible to succeed with any platform — through Profiles does particularly well, owning 6 of the top 10 spots.

  • Profiles average score = 56%
  • Elsevier Pure average score = 42%
  • Custom software average score = 37%
  • Vivo average score = 31%
  • SciVal Experts average score = 22%

But this is complicated by the fact that systems that are hosted on a domain name other than that of their parent institution perform much worse:

  • Official domain? (e.g. vivo.cornell.edu)
    average score = 49%
  • Other domain? (e.g. experts.scival.com/asu)
    average score = 21%

Putting that together:

  • SciVal Experts + official domain average score = 59%
  • Profiles + official domain   average score = 58%
  • Elsevier Pure + other domain average score = 42%
  • VIVO + official domain average score = 38%
  • Custom + official domain average score = 37%
  • Profiles + other domain average score = 35%
  • VIVO other domain average score = 18%
  • SciVal Experts + other domain average score = 14%

While the n for these are small, it gives us a sense of which combinations may work best. I was particularly disappointed to see that VIVO sites don’t score very well, on average, compared to other platforms.

The final lesson learned is how important it is to get a high volume and diversity of incoming links to help increase search rankings for a given site, something made clear by the Moz Search Engine Ranking Factors.

The top 3 sites have a huge diversity of incoming links:

  1. findanexpert.unimelb.edu.au has 488 linking root domains
  2. profiles.umassmed.edu has 249 linking root domains
  3. profiles.ucsf.edu has 858 linking root domains

The correlation between linking root domains and search rankings holds true across our dataset:

linking root domains

# linking root domains via Moz Open Site Explorer. Data excludes experts.scival.com.

 

4. The five steps to increase your rankings

Already have a site running on an institutional domain? Here’s how to move ahead:

1. Be worthy of being linked to:

  • Most people care about people, not generic information-finding site
  • Make profile pages beautiful and chock-full of information, so people will want to link to them

2. Establish benchmarks:

  • Install Google Analytics (or equivalent) on every page
  • Learn how to use it — we also recommend Web Analytics 2.0 by Avinash Kaushik

3. Get good with Google:

  • Add a sitemap.xml (sitemaps.org)
  • Register on Google Webmaster Tools to:
    • register your sitemap
    • catch indexing errors early
    • link to your Google Analytics account

4. Make sure your pages will look their best on search engines:

anatomy of a search result

 

5. Kickstart the process of getting linked to:

  • Get campus sites to link to your homepageas a trusted campus resource
  • Get campus sites to link to individual profiles from departmental profiles, news stories, directory, etc.
  • Encourage reuse of your data via APIs, and ask for a link back as attribution

If that works:

  • Researchers will start linking to their profile pages on their own sites
  • Blogs and social media will start linking to your profile pages as authoritative sources
  • Departments may start linking to your profiles if your data is more current than theirs

2 thoughts on “RNS SEO: How 52 research networking sites perform on Google, and what that tells us

  1. Thanks, Anirvan! As always, great analysis with telling illustrations and clear, doable recipes for improvement! One thing I wonder about is the link between the software choice and the SEO performance. Aside from the hosting domain and the structure of profile URLs, what other factors built in the software may affect SEO? If the recommended strategy (five steps to improve) is implemented equally well on two RNS using different platforms, would their SEO outcomes still differ?

    • The best guide I know is the Moz 2015 Search Engine Ranking Factors study, particularly the list of correlations. For me, at a very high level, the three main takeaways are:

      1. Get a diversity of relevant (and organic) links
      2. Offer relevant content that engages users and loads fast
      3. Be a good web citizen — link out broadly, use reasonably semantic markup, etc.

      But there are so many individual variables (Moz looked at 170+ factors), it’s hard to imagine that any two sites could get similar results, even if they tried implementing similar strategies.

      I think software choice matters because software hardcodes in several defaults (e.g. URLs, link structure, page markup, etc.) — but the hard work of evangelizing a site is where SEO performance can live or die. Having the perfect software doesn’t matter if nobody thinks your site is worth linking to.

Comments are closed.