AARP launched a pilot program for member-exclusive premium content that was more in depth and comprehensive than other content on the web site. The premium content was collected in an online library on the AARP web site. The content included PDF files, HTML articles, audio, and video.

The landing page for the library provided an index of the all the premium content organized by categories such as health, money, entertainment. As new content was created, the landing page was reorganized, so content placements on the page varied from week to week.

To create a forecast, data from past landing page click-throughs was modeled through a linear regression analysis that incorporated location on page, content topic and format. A regression model was created for desktop and phone landing page layouts.

After creating the models, it was possible to forecast click-throughs for the various pieces of content and compare against actual click-throughs. A visualization was created to capture the over and under performing content.