Last June, jQuery Foundation held their conference in beautiful Portland, Oregon. As a Diamond Sponsor, we wanted to build something that would be beneficial to the jQuery community part of our Splunk4Good initiatives. What’s better than Splunking the entire conference?
To see the end result, check out this interactive infographic showcasing Splunk-powered web analytics applied to the conference website. The complete Splunk dashboard can be found here.
The goal is to capture client-side data (e.g. pageviews, link/button clicks, hovers), and build powerful analytics & visualizations in order to tackle the following business questions:
- Which topics are visitors most interested in?
- What are the top traffic sources for visitors who purchase tickets?
- How are visitors interacting with the site, including time leading to ticket purchase?
To help answer these questions, we collected a sizeable clickstream during the week of the conference:
- 8,500+ unique visitors with 15,000+ pageviews
- 240,000+ client-side events including 60,000+ clicks
1. Which topics are visitors most interested in?
Every time a visitor expanded a talk to read its description, it was recorded as a click. The underlying assumption is that the talks with the most clicks, are the talks that are most intriguing to conference attendees, and therefore tend to be the talks that are most popular.
Actionable insights:
- Organizers can ensure future conference programs include highly targeted content by focusing on most popular tracks, as listed above.
- Organizers can more optimally plan room capacity based on a talk’s anticipated popularity. For example, “jQuery UI Widgets vs HTML5” by TJ was a front-runner early on before the conference started; with that predictive knowledge we could prevent major room overflows.
2. What are the top traffic sources for visitors who purchase tickets?
While jQuery conference site is primarily for content consumption, it does act as a store front (yes, they sell tickets!). So we tracked the different steps (initial pageview, click on ‘buy’ button, and order confirmation) that led to ticket sales, and broke down the funnel by traffic source.
Actionable insights:
- Surprisingly, a local site, calagator.org, had the highest conversion ratio of 3.85% or 8x more than jquery’s own blog. It turns out that site caters to the niche technology community of Portland. Therefore, conference marketers should consider featuring events on local community portals to capitalize on such high conversion sources.
- Conversions from blog.jquery.com are about 2x more than the rest of the jquery properties like jquery.com, api.jquery.com and events.jquery.org. Thus conference marketers may want to further leverage jQuery blog & increase cross-promotion to market their event, more so than other jQuery subdomains.
3. How are visitors interacting with the site?
Finally, as with most sites, a lot of insight can be drawn by having a view of visitors’ timeline of actions: this is similar to Google Analytics Visitors Flow except it’s at the more granular event level vs. page level. Each visitor has its own ‘swimlane’ composed of a series of shapes: rectangles depict time spent on a specific section, and circles depict specific actions such as clicks. We like to think of this as the Visitors Behavior Flow:
Actionable insights:
- Using the Visitors Behavior Flow, you can monitor user behavior on the site in real-time in order to detect potential user experience problems early on, or even provide targeted on-time support or promotions…etc.
- While this sample data is small, you can visually observe that people who actually purchased tickets spent about the same amount of time on both the training and program sections, yet the training section is at the bottom of the page. This could mean training classes are at least as important as the session talks to this audience. Organizers may want to focus on the training section just as much as the program itself to further drive attendance and ticket sales.
Now it’s your turn to create your own powerful data visualizations on Splunk!
While Splunk offers easy-to-create dashboard visualizations, this infographic was built exclusively using open source JavaScript libraries such as jQuery, Backbone.js, Require.js, and D3.js – tools web developers are already familiar with – in order to show how one can fully customize the way Splunk search results are rendered.
Email me (Roy Arsan) or Shirley Wu and share with us your own innovative ways of visualizing your Splunk data to better understand your users.
Stay tuned for technical details on how to track & collect data from your own sites, or join us at our Digital Intelligence booth at Splunk .conf 2013 next week in Las Vegas:
Your Data, No Limits!