by Sara Ring
Quick Summary
My colleagues and I attended the November Amigos virtual conference Innovating with Metadata. The conference opened with a keynote address by Amanda Sprochi about the new International Federation of Library Associations (IFLA) Library Reference Model (LRM) and how it will have a major impact on RDA (Resource Description and Access). While there were several breakout sessions to follow, I would like to summarize one in particular: Analyzing Linked Data From and SEO Perspective. For several years I’ve been following developments in library linked data. This is one of the first speakers I’ve heard that analyzed library linked data from a search engine optimization perspective.
My colleagues and I attended the November Amigos virtual conference Innovating with Metadata. The conference opened with a keynote address by Amanda Sprochi about the new International Federation of Library Associations (IFLA) Library Reference Model (LRM) and how it will have a major impact on RDA (Resource Description and Access). While there were several breakout sessions to follow, I would like to summarize one in particular: Analyzing Linked Data From and SEO Perspective. For several years I’ve been following developments in library linked data. This is one of the first speakers I’ve heard that analyzed library linked data from a search engine optimization perspective. Trey Gordner, (founder of the startup Koios) noted that while one of the promises of linked data for libraries is search engine visibility, few quantitative studies have been done to analyze linked data's impact on search engine rankings.
Gordner chose to analyze Zepheira’s Library.Link Network project, for a few reasons. It is one of the most well-known projects, they are experts in the field, and the claim is that publishing library data as linked data will lead to increased visibility on the web. The presenter stated that there are some problems that can be fixed, but some are larger conceptual problems with library linked data, potentially shared by all library linked data projects.
He showed us the structured data on the webpage of the book A Separate Peace from (Delaware County District Library) as an example throughout the rest of the session. Taking us through each category from Google’s SEO Starter Guide, he evaluated and assigned grades (A-F) based on how well the page followed Google's search engine optimization best practices.
For example, here’s a snippet of the structured data from the book A Separate Peace from Delaware County District Library’s Library.Link page.
<title>A Separate Peace - Delaware County District Library</title>
<meta property=”og:descrption” content=”An American classic and great bestseller for over thirty years...”/>
Looking at the first section of the Google SEO Starter Guide, SEO Basics, it offers some best practices for webpage titles and descriptions. Here were the grades assigned after analyzing the snippet of structured data above.
- Titles are accurate, Grade = A
- Titles are unique, Grade = C
- Titles are brief (< 60), Grade = B
- Descriptions are accurate, Grade = A
- Descriptions are unique, Grade = C
Here’s a summary that shows the rest of the SEO categories, the grade the presenter applied based on analyzing the book library linked data example, and a summary of best practices according to Google's SEO guide.
SEO Basics, Grade = B
Add appropriate meta description tags, ensure uniqueness
Site Structure, Grade = F
Avoid duplicate pages, no-crawl technical pages (e.g. authority), use a hierarchy structure, create sitemaps
Content, Grade = C-
Find opportunities to be unique, design for users (not bots)
Linking, Grade = D
Build natural inbound links, no-follow internal links that don't help the user
Engagement, Grade = ?
Prioritize user experience, monitor web metrics (looking more at this one in January 2018 Code4Lib Journal article)
In summary, his interpretation was that there is a reason for concern. More recently Google penalizes link schemes (which library linked data can look like) and prioritizes engagement. The more libraries that publish their bibliographic data as linked data, the more duplicates there are on the web. “Linked Data is not SEO” was the big takeaway from this session, and publishing your library data as linked data does not ensure your data will show up in search results. He encouraged libraries who wish to or plan to publish their data as linked data for greater search engine visibility to consult with an SEO professional.
Trey Gordner will have a Code4Lib Journal article coming in January 2018 that will look at several library linked data projects from a search engine optimization perspective.
