Recently we implemented Google Analytics to track usage of our digital collections.  Sean has already contributed several great posts about our digital collections use statistics, but one thing I find particularly interesting (and amusing) is that Google Analytics allows us to see the types of keywords our users are entering into Google, Yahoo, and other search engines, and where those keywords lead them in our digital collections.

Not surprisingly, some search queries are common and reveal the subject strengths of our digital collections.  For example, the top three queries that bring users to our collections are “sheet music,” “ad access,” and “history of advertising.”

After scanning through thousands of these search queries, several distinct categories emerge: the known-item query (an exact title in quotes), the URL as query (e.g.  http://library.duke.edu/digitalcollections/adaccess/), and the format query (e.g. “diaries” or “manuscripts”), among others.  The most entertaining category, however, is the query issued in the form of a question.

Below are some of the important questions our users have asked with links to where they’ve found answers to those questions in our digital collections.

/* Style Definitions */
table.MsoNormalTable
{mso-style-name:”Table Normal”;
mso-tstyle-rowband-size:0;
mso-tstyle-colband-size:0;
mso-style-noshow:yes;
mso-style-parent:””;
mso-padding-alt:0in 5.4pt 0in 5.4pt;
mso-para-margin:0in;
mso-para-margin-bottom:.0001pt;
mso-pagination:widow-orphan;
font-size:10.0pt;
font-family:”Times New Roman”;
mso-ansi-language:#0400;
mso-fareast-language:#0400;
mso-bidi-language:#0400;}

• “what is electronics?”

• “can you rent cars at Hertz for fun?”

• “what to wear with a corset”

• “what is a funeral chair?”

• “what are 1955 war bonds worth today”

• “what are chemical properties of Listerine?”

• “will I be sick tomorrow?”

As Tom mentioned, we’re in the process of re-examining our homepage (and the layout of individual collection homepages).

What do people actually do when they come to the Duke Digital Collections homepage as it is now?  One way to tell is to review our server logs.

Here’s a look at the year in review.

2008 At a Glance:  68,325 homepage hits

Continue reading Home(page) Economics →

Looking back at our 2008 web logs, we can learn a lot about how our system and our collections are being used. We hope to combine an analysis of this usage data with usability testing and other modes of evaluation to better inform our continued development of our system & interface in 2009.

Here are two separate charts (below): one for the first half of 2008 (Jan – June) and the other for the second half (July – Dec). The one on the right includes more collections (we introduced several throughout the year) and may be a more representative look at the usage.  Also keep in mind that the collections vary in size (larger collections have more items *to be viewed* and often have more ways to formulate queries).

Click to enlarge:

Jan – June 2008

July – Dec 2008

Interesting findings

Continue reading Collection Usage Stats for 2008 →

Now that 2008 is over, we’ll be posting a few charts & graphs that illustrate some interesting trends in how our digital collections (and our shiny new system) have been used in the past year. This post focuses on “referrers,” or, those other sites that people come from that directly lead them to land on our pages.

OK, so what are we counting?

• Any referrals from outside of Duke Libraries websites that resulted in a Page View anywhere in library.duke.edu/digitalcollections/*, including our digital collections homepage, individual collection homepages, contextual info pages, item pages, or search results.

How many?

• 890,000 referrals from 10,000 unique external domains (all Duke library web sites/pages excluded). Only the top 9 individually account for more than 1% of external referrals, so there’s quite a long tail.

Notable External Referrers

Of the 10,000, some stand out in particular…
Continue reading How We’re Found (or, Referrer Stats for 2008) →