MEASURES AND METRICS
DOMAIN: ONLINE INTERACTIONS
Most Repositories maintain a website, and many offer searchable databases of collection holdings and digital content, as well as a presence on one or more social media services. The online interactions domain includes several of the more common measures and metrics that a Repository can use to assess how frequently Users are accessing and interacting with content the Repository posts in the online environment. Online interaction statistics can be obtained by web analytics tools and social media services. Such tools and services generally offer many more types of useful measures and metrics than can be described here.
Basic measure (“Page Views”)
Count the total number of Page Views by Users of online content posted by the Repository. Page Views may include, but may not be limited to, the Repository’s website, finding aids, online exhibitions, and digital collections.
Rationale:
Page Views are the most common as well as the most basic measure of User activity on a website. All web analytics tools are able to tabulate the number of Page Views that a website receives during a determinate period of time, and to do so in a manner that is consistent with other such tools,whether free or commercial. Page Views therefore offer an easy and useful index for comparing website activity at single institution across different time periods, or across multiple institutions over the same period.
Guidelines for collection:
- Select and configure an appropriate web analytics tool to collect data from the desired websites managed by the Repository.
- Combine all Page Views for all monitored websites to obtain the basic measure.
- If possible, exclude Page Views by the Repository’s own staff (for example, by filtering out IP address ranges assigned to staff workstations).
- Exclude statistics derived from social media accounts maintained by the Repository. See social media reach advanced measure.
Applications and Examples:
- A Repository configures Google Analytics to analyze traffic on its main websites, its online exhibitions, and its finding aid database. During a given month, the Google Analytics reports 6,765 Page Views on the main website, 3,444 page views across all online exhibitions, and 940 Page Views of its online finding aids. Count 11,149 Page Views.
Advanced measure (“Unique Page Views”)
Count the total number of unique Page Views by Users of online content posted by the Repository.
Rationale:
Unique Page Views offers a more a precise measure of User activity on a website than Page Views because it filters out multiple views of the same page by the same user during the same session/visit.
Guidelines for collection:
- Ensure that the web analytics tool selected by the Repository can report unique Page Views. If a given web page is viewed multiple times by the same user during same session/visit, it is counted as one unique Page View.
- If possible, exclude the unique Page Views of Repository staff (for example, by filtering out a block of IP address ranges es assigned to staff workstations).
Applications and Examples:
- A historical society configures Piwik to analyze traffic on its main website and its digital collections website. During a given month, Piwik reports 4,501 unique Page Views on its main website and 2,934 unique Page Views for its digital collections. Count 7,435 unique Page Views.
Advanced measure (“Sessions”)
Calculate the total number of sessions initiated on websites maintained by the Repository.
Rationale:
The number of sessions initiated on websites maintained by a Repository provide an indication of the number of Users who are visiting those websites. Since the same User may visit the websites from different devices, web analytics tools have moved away from describing the activities of users and visitors to describing instead the characteristics of sessions.
Guidelines for collection:
- Ensure that the web analytics tool selected by the Repository can report website activity in terms of sessions.
- A session may be defined as group of interactions that take place on a website within a given time frame. A session may include one or Page Views, file downloads, and other types of events.
Applications and Examples:
- A researcher visits the main website of an archives from a computer at home to check Reading Room hours and policies. While traveling to the archives, the researcher visits the website to check the address and directions. Since different devices are used to visit the website, the Repository’s web analytics tool should count 2 sessions.
Advanced measure (“Session Duration”)
Calculate the total amount of time a User spends viewing a website during a single, continuous viewing period.
Rationale:
Session duration (or length) is a measure of the cumulative amount of time a User spends on the website during a single, continuous viewing period, generally termed a session or a visit. Session duration includes time spent viewing and navigating between pages within the same website. Longer Session durations indicate greater User engagement with site content.
Guidelines for collection:
- Ensure that the web analytics tool selected by the Repository can report Session duration.
- Check and adjust, if desired and available, the configuration setting for session expiration time. For most web analytics tools, a session expires after 30 minutes of user inactivity.
Applications and Examples:
- A User views an online Exhibition and stays active on the page for 15 minutes, becomes inactive for 10 minutes, and then becomes active again for another 10 minutes. The total session duration for that User would be 35 minutes.
Advanced measure (“Downloads”)
Count the number of files downloaded from websites managed by the Repository.
Rationale:
For Repositories that post content online for Users to download and store locally for immediate or later use, tracking file downloads can be an effective way of understanding which types of content Users value most and which files are most popular. Files that can be posted for Users to download may include PDF versions of finding aids or oral history transcripts, digitized photographs and other digital images, audio and video recordings. Counting file downloads can provide a more precise means of analyzing User interest in a Repository’s holdings than website Page Views.
Guidelines for collection:
- Ensure that the web analytics tool selected by the Repository can report file downloads. Some tools treat the downloading of files as events.
- Statistics for streaming media files should be excluded from file downloads, but Repositories that post streaming media online may wish to track usage statistics for those media separately.
Application and examples:
- If a high-resolution image is embedded in a web page, a web analytics tool will treat User’s visit to the page as a Page View. If a web page includes a thumbnail version of the image which a User may click to download and view a high-resolution version of the image, the action of clicking on the image would be treated by a web analytics tool as a file download.
Advanced measure (“Download File Type”)
Define categories of digital object types or file formats for tracking, and count the total times each object or file type is downloaded.
Rationale:
Knowing which file types are most often downloaded may help a Repository determine which types of downloadable files its Users prefer and to prioritize its creation of digital content accordingly.
Guidelines for collection:
- Ensure that the web analytics tool selected by the Repository can identify and report the respective numbers of file types of downloaded files the Repository wishes to track.
- File formats may include PDF, Word, TIFF, JPEG, MPEG, MP3, WAV, etc.
- Categories may include, but are not limited to textual, image, moving image/video, and audio formats.
Application and examples:
- Reviewing its web analytics report, a historical society discovers that 90% of the file downloads from its oral history collection were .mp3 files and only 10% were .wav files. Since .wav files are much larger and require more bandwidth to deliver, the historical society decides to discontinue offering Users the option to download .wav files, providing them instead in response to specific requests and and otherwise keeping them preservation purposes.
Advanced measure (“Traffic Source”)
Determine and count the sources from which online traffic is directed to the Repository’s website.
Rationale:
Determining and counting how users are directed to a Repository’s website can help the Repository understand how Users are discovering it and what opportunities it may offer to improve its marketing outreach. Incoming traffic sources could point to successful outreach initiatives involving other websites such as adding a link from a Wikipedia entry to related digitized material or a finding aid.
Guidelines for collection:
- Ensure that the web analytics tool selected by the Repository can report traffic sources.
- Some web analytics tools can distinguish traffic sources by type, such as direct traffic (from a typed url, bookmark, message link), referral traffic (from links in another website), and search engine traffic (from search engine results).
- Traffic sources can be analyzed for a Repository’s overall web presence or focused on particular sections of a website, such as online Exhibitions.
Application and examples:
- An archives discovers that 10% more of its website traffic has been coming from Wikipedia following an edit-a-thon workshop during which Repository staff systematically added links to its online finding aids from several Wikipedia entries.
Advanced measure (“Social Media Reach”)
Count the total number of interactions with the Repository’s social media services.
Rationale:
Social media services generally offer one or more methods for Users to leave an intentional trace of their engagement with content presented through the service by clicking a “like,” “follow,” “share,” “repost,” or “comment” button or link. Counting the aggregate numbers of such interactions can enable a Repository to gauge how popular it is among the social media audiences it attempts to reach.
Guidelines for collection:
- Count the number of interactions within each social media account maintained by the Repository, categorizing interactions by type (e.g., “likes,” “comments,” etc.).
- In addition to aggregating counts from different services to obtain an overall measure of social media reach, Repositories may wish to track the counts for each social media service separately in order to facilitate longitudinal comparisons and evaluations of the distinct reach and impact of each.
Application and examples:
- A Repository that maintains a Facebook page and two or more Tumblr accounts may wish to tally separately the number of Facebook likes and shares while tallying and combining the total number of likes and reblogs across all Tumblr accounts.
- A Repository that maintains a Twitter account may wish to count the number of tweets that it publishes as well as retweets and likes of its own tweets.
Recommended metrics
Total Page Views per day/week/month/year
- The total number of Page Views over a given period of time can provide an indication of the number of visitors and visits to a Repository’s website, although it more directly indicates the overall volume of website activity.
Total sessions per day/week/month/year
- The total number of sessions over a given period of time provides an indication of frequently a Repository’s website is visited by users.
Total session duration per day/week/month/year
- Analyzing the total duration of all sessions during a given time period and comparing it against totals for the same intervals at other times can provide a complementary measure of User interest in a Repository’s website content to total Page Views.
Average session duration per day/week/month/year
- Monitoring the average length of time Users spend on a Repository’s website can serve as an index of User engagement with the website’s content. Some analysts have suggested that Users who spend more than 60 seconds on a website have demonstrated interest and commitment.
- Web analytics tools generally calculate average session duration (or length) by dividing the total duration of all sessions by the number of sessions.
Page Views per session
- Analyzing the numbers of Page Views (or pages visited) per session provides an index of user engagement with the Repository’s website. Users who are more engaged with a website tend to move from one page to another in a website rather than leaving the website after viewing only one page (i.e., “bouncing”; “bounce rate” is another useful metric that many web analytics tools can provide).
Total file downloads per day/week/month/year
- Calculating the total number of files downloaded from a Repository’s website during a given time period and comparing it against totals for the same intervals at other times can provide a complementary measure of User interest in website content to Page Views and Session duration.
Average number of files downloaded per session
- Calculating and monitoring the average number of files downloaded per session can provide an indication of how Users are engaging with the Repository’s website. Are users coming to the website with the expectation of being able to download content?
Table of Contents
Measures and Metrics:
