I've got questions
Home | Blog | usability | Thinking about web analytics

Thinking about web analytics

(Today’s post is almost more of a question than anything else, so please post your thoughts in the comments.  I’m genuinely interested.)

Tape measureI’ve got questions.

Most libraries have some form of web analytics, to measure at least the basics:  page views, number of unique visitors, time spent on the site, and so forth.  But those numbers are often largely meaningless.  Sure, maybe your library’s site got more page views than last month, but so what?  What does that translate into?  More views don’t necessarily mean that your library will have more (for example) circs or more attendance at programs.

The relevant term here, used in web analytics,  is conversions.  A conversion is when a visitor comes to a website and then takes whatever action the owner of the website is hoping the visitor will take.  For retail sites, defining a conversion is generally easy.  Someone came to the site and then bought something.  Even for other not-for-profits, figuring a conversion is relatively simple.  Typically, they want you to donate money or take an advocacy-related action, such as writing your legislator.  In both of these instances, there are clear actions that the site owners want visitors to take.  The conversion is, essentially, the holy grail of web metrics.  It’s the thing that really matters.  Did the website help users take the desired action or not?

Now, back to libraries.

To measure a conversion, you have to first have at least one goal for site users.  I’ve often seen goals like “helping users find what they want to do more easily.”  Sure, that’s great.  But how do you convert that to something the library wants users to do?  Libraries often want to  do things like circulate more items or increase program attendance.  Both are offline actions.  How can web analytics be connected to offline goals?

Metrics can help libraries tell stories, target specific audiences and better arrange their websites for users.  In no way am I saying that metrics are unimportant.  ( If you don’t have them, get them.)  But, I’m truly curious about how libraries figure conversions, when conversions measure goals met.  In other words, how can a library really calculate success via its website?  The website is primarily an outreach function of the library, not the library itself.  Metrics might tell you how easy to use or interesting your library’s site is, but I wonder how well they can tell us how successful the library is.

What am I missing?  Folks who are heavily steeped in analytics–can you provide any insight?

There are 4 comments

  1. 1) My local library has the Gutenberg collection mirrored, as well as several of the Baen CD-ROMs and a number of other free book collections. The Louisiana State Library allows them to count hits to these books as circulations.

    2) A patron survey might give you an idea of how often people are coming to events such as storytime or author signings because of visits to your website. This, combined with the analytics, especially dwell time, might be able to give you some idea of which parts of the site are attracting the most attention. Dwell time is a really hard thing to measure, other than when the viewer uses a link on a page to navigate to another page on your site (unless you’re using redirector links to force a hit on your site when they click to move away from you – worth considering, even if I personally hate them…)

  2. Page views, visitor numbers, and time spent are just the tip of the iceberg of what you can do with analytics. I can’t speak for all analytics products but Google Analytics lets you delve much deeper. Just a few of the things you can do are:

    1. See in-page clicks that tell you what links on a given page are most used
    2. Visitor flow that shows you page by page navigation of how people move through your site
    3. Social media metrics
    4. The ability to set up goals that track metrics you define so you can see how successful a new piece of content (like an ad or service) is.
    5. Learn what search terms people are using to find your site and what terms they’re searching for within your website.
    6. Learn the number of people using mobile devices to access your site, including which devices they’re using.

    You get out of analytics what you put into it. So while analytics can’t tell you everything about your users when paired with usability testing and more qualitative methods of getting to know users, like chatting at the reference desk, it can be extremely helpful to get an idea of what people are using your website for.

    And I would also argue that a library’s website is not just an outreach tool. The library website -is- the library, just online. On most sites people can do everything they’d be able to do coming in to a building, sans picking up physical media. They can check out books, get help from a librarian, get digital media, look up journals, do genealogy research, learn a language, and in some cases view library programming. It can and should be used for marketing and promotion but these days social media does that just as well, while the website is a resource. The library website is a digital branch.

    Circulating more items or trying to increase attendance at a program might be at the end offline (though that can be argued for circulation stats these day with ebooks) but the beginnings can be on the website. Analytics can be used to see how different methods of promotion on a site, as well as the site architecture itself, can help or hinder the goals you want your users to complete. It’s certainly not the end all of information gathering but it can be an extremely powerful tool in the box.

  3. I am guessing my post was not as clear as I was hoping. What I really want to know is if any libraries are figuring true conversions: being able to pinpoint how online behavior translates into (or doesn’t) goals that are achieved *offline.*

  4. I’ve tried to point out to one of my librarian friends that metrics and analytics of *offline* patron behavior can tell libraries the same type of useful info. For keeping track of reference desk usage, they currently use – no kidding – a pencil and paper with tick marks ( you know, IIII is four and to make five, you put a slash mark thru that … ) and then someone has to total that up every afternoon, and THEN hand-jam the numbers into a spreadsheet for the “monthly” … sigh … put a laptop right there at the reference desk, already.

Comments are closed.