We took a two-pronged approach to auditing OPL's site content.

First, we examined the basic pages. OPL has around 200 pages on its website. For each page, we assigned an ID number, then gathered:

• Basic identifying information. Navigation, Title, URL, Content Type. We included a column (N) for Metadata, but found that none was applied to any basic pages.
  In some cases, we divided pages into "chunks" of content so it can be split into multiple parts when migrated.

• Summary. Topic Area (our quick attempt to categorize the content), Description, Main Message (like a headline for the content).

• Audience information. Who is the intended audience? We used Hemingway to determine a reading level for the text.

• Files. Images, Image Sizes, Attachments.

• Analytics, covering a time period of January 1 2019 - January 31 2021. (If the page was created after January 1 2019, we used the entire time the post was live.) We collected total views and average minutes spent on each page. We then calculated average views per month, since that number is a smaller and a little easier to grasp, and also would allow us to compare posts that had been active for different time periods. Under Analytics Notes, we indicate any special circumstances for the data we gathered, such as whether there were any months of unusual activity that skewed the data.

• Notes, Type, Action. Our assessment of the page, its contents, and best strategy for migration (or purging!).

Second: blog posts.

OPL staff have maintained various blogs since 2013. At the time of our assessment, there were over 1,200 blog posts on OPL's website. They are original pieces of content created by staff, and they vary tremendously in readability, continued relevance, alignment with OPL's mission statement, and resonance with readers. OPL doesn't want to migrate eight years of blogs that may or may not have value, but they don't want to wipe the slate clean, either. They want to use what's useful and has been enjoyed by their readers - and they want to produce more content like that, and less of the content that has failed to attract attention.

Using Google Analytics, we exported a complete list of OPL blog posts, then cleaned and sorted the data in Excel. We looked at views over the last two years, since that's what we'd used for basic pages, and that's quite a long time in the blog world. For each of OPL's four most-used blogs, we calculated the average number of views on a post over the last two years,* then highlighted posts that had been viewed more than twice as many times as average. That resulted in about 12-25 posts, which is within the range OPL is willing to migrate.

Once we had those most popular blog posts, we began assigning tags and looking for common categories. We identified a few topics that seem ripe for OPL to develop future blog content and publish pages: Oakland history, African-Americans in the Bay Area, explaining library services, "library from home," book recommendations, and school.

Content matching those topics can be migrated and mapped to those pages. OPL can recruit bloggers to write specifically and only on those topics. Other blog content can be quietly retired.

*For content that was less than two years old, we calculated the number of views from the date posted through January 31, 2021.