LavaCon 2012-Project Metrics and Development Team Management Track
Presenter: Bob Zebian, IBM
Bob Zebian brought a real life example of web analytics value to documentation teams to his LavaCon 2012 session. He offered the following case study on how analytics at Sterling Commerce helped reinforce movement from a book paradigm to an online paradigm. He briefly covered the two types of analytic tools available: log analysis tools and script-based analytic services.
Some of the standard web analytics useful to documentation:
- Visitors, visit
- Pages viewed
- Least traffic pages
- Viewing duration
- Navigation path
- Search keywods
- User location
- Browser/OS/Mobile device type
- File downloads
Statistics at Sterling Commerce showed that users were not reading documentation all the way through, and only looked at an average of 5 pages per visit, with an average of about 5 minutes on site. Also, 45% of users were using search to find content. Analytics can save money in areas other than documentation, by revealing the commonly used and viewed items, so resources can be focused on testing those items in quality assurance. It can also show development/product management that there is low usage of a particular browser, and that the cost justification is not there to support that browser. Other analytic uses can include reduce customer cases by identifying high-volume cases that can be addressed by better documentation, analyze call volume pre- and post-project, and quantify the results in monetary form. You can also improve search by analyzing the keywords used by customers.
If you want to get started using analytics to provide value, here are some of the questions you can provide statistics to product management:
- What are users looking at?
- What aren’t users looking, and why?
- Where should we invest our time?
- What types of issues are driving our customers to the documentation?
- Are customers reviewing hot fix/maintenance releases?
- Working with support, consider how web analytics data aligns with support calls and open tickets?
- Working with product manager, how do user profiles impact product direction?
When you look at the tool options, make sure to check with your marketing department first. You may be able to piggyback on a corporate web team analytics service, and just build reports that give you the information you need. There are also free options, such as Google Analytics, or paid services that are hosted by vendors, such as MadCap Software’s Feedback Server. One of the most serious considerations when looking at tools is how much you need to modify your content to accommodate the solution. For instance, if you already work in MadCap Flare, adding Feedback Server usually requires some backend IT work, and adding some script in a template, but doesn’t require you to rework your current content to be analyzed. This is one of those situations where it’s better to involve other departments.
Zebian then covered what you can do with the data, including downloading the data into a spreadsheet and using Excel functions such as text-to-columns, concatenate, and subtotal to slice and dice. He stressed that documentation teams should be aware of privacy and legal concerns. Most analytics tools use cookies and JavaScript to report user activity. If you use analytic tools, make sure your company’s web privacy policy clearly states the use of cookies and that information is gathered. Also, you need to be aware that some countries require instructions to allow users to opt out of tracking.
This session was very useful to me, as our department is getting ready to choose an analytics tool for our product help. Zebian provided a good grounding that I can take back to my manager to make sure we’re gathering the right information and addressing all the concerns.
Slideshare link: http://www.slideshare.net/bzebian/lavacon-2012-how-documentation-teams-can-use-web-analytics-to-expand-their-corporate-value
