This year should mark the beginning of a remarkable recovery in the world economy, and I find myself extremely fortunate to be counted amongst the hundreds of attendees at eMetrics Toronto 2010.  Having attended the show in 2008 but not last year, I witnessed a significant evolution and have to admit the folks at Rising Media have done a fantastic job of transforming what was once a much smaller, admittedly lack-luster conference into a major, must-see event.  Perhaps this is due to a natural maturation process, but I’d like to attribute this success to the increasing awareness and respect the online business community has amassed for web analysis.  The resounding theme of the show has evolved as well.  What once was “how do we do X” has quickly become “define goals, provide actionable insight, measure results”.  Whether through quantitative or qualitative means, the focus has become listening to the voice of the consumer.

Define Goals

Many panelists and speakers admitted that their organizations are still mired in poorly defined goals, however they failed to respect that the mission of a web analyst or business analyst is an evolutionary approach.  Avinash Kaushik explained it best when he described a “virtuous data quality cycle” in his first book Web Analytics: An Hour a Day.  The point is, you will have overall business impact goals, and ad hoc goals that spring to life, die off, change into other objectives, and that’s okay.  Some of the speakers and panelists said it best:

  • Theresa Locklear, NHL.com:
    • Goals for analysts were clear, they wanted people to visit the video player more often, change channels more often, and share video with their friends.
    • Business goal is a bit veiled, but it is clearly to generate additional sales and ad revenue.
  • Avinash Kaushik, Google:
    • “Most companies think social media is about shouting, this is a dumb ass strategy”
  • Theresa Kushner, Cisco:
    • Our vision is planned 5-8 years out, strategy: 2-4 years, execution: 18 months.

Provide Actionable Insights

Great, we have all this data, but what does it mean?  At a minimum, what should distinct groups or areas of the website do differently? At a macro perspective, what are visitor behaviors and voice of customer telling us about what the business should do differently?  Here are some great nuggets of information mentioned at eMetrics about actionable insights:

  • Avinash Kaushik, Google:
    • “Focus on outcomes, find out the ‘WHY’ in addition to the ‘WHAT'”
    • Put data in context, “context means comparison of both internal metrics and competitive intelligence”
    • Many first time visitors to landing pages don’t convert, so stop screaming at them, soften messages on PPC landing pages and build a better initial experience.
  • Theresa Locklear, NHL.com:
    • Highlights are key, people love highlights so show more highlights
    • The Hockey Show isn’t generating much engagement, is costly to produce, and is a longer production piece which means less opportunity to run pre-roll advertising, consider dropping it
  • Michael Helbling, Land’s End:
    • We look for smart lazy people, because smart lazy people automate things so they have time for analysis.
  • Shari Cleary, MTV:
    • Don’t just send data, ask the question WHY and find the specific business requirements behind any request for data.

Measure Results

How can you determine whether your defined goals were met, or whether you were successful or not?  Results must be measured and compared to historical data or competitive intelligence.  What can you measure? Pretty much everything, so the question becomes what SHOULD you measure:

  • Avinash Kaushik, Google:
    • “There is more data in tools than God wants you to have”
    • “I like Klout because it gives follower retweet %”
    • “I like analyzewords.com, every word you write gives you a window into your cognitive world, scary crap!”
    • RE: Paid analytics tools being very similar to free tools: “Flushing money down the toilet makes a cool sound”
    • “Don’t ignore the long tail or search, it can account for more than 7 times the visits versus top 10 keywords”
    • RE: Content strategy, “Consider % content per section versus % of visits, create what people wants”
  • Theresa Locklear, NHL.com:
    • Use multiple data sources and categorize by source participation. First step is segmentation, second step is persona development (see pyramid graphic on the right)
    • “We considered internal production costs versus sell-through ROI when evaluating video content”
    • Analytics insights drive almost everything we do now, editorial content, PR functions, website development, video production, etc.
  • Theresa Kushner, Cisco:
    • Be sure you know what you want to measure, gain insight through analysis, share those insights, encourage marketers to use that information.
  • Shari Cleary, MTV:
    • There is a clear expectation from multiplatform males that they expect stuff like South Park on every screen/platform. Fueling a vertical strategy.
Be Sociable, Share!