How do you know if you've been successful with search engines and your website in general? You can check your "rank" at search engines for particular keywords, analyze log files to see the actual terms people used to reach your website, or make the ultimate jump and "close the loop" by measuring sales conversions and ROI. This panel explores both classic and cutting-edge techniques to measure success, what statistics you should really care about, ways to be more strategically focused, and how to drive increased revenue for your business.
Moderator: Richard Zwicky, Founder & CEO, Enquisite
Speakers: Jim Sterne, Target Marketing & Chairman, Web Analytics Association Matthew Bailey, President, SiteLogic Avinash Kaushik, Author, Blogger, Analytics Evangelist, Google Marshall Sponder, Senior Web Analyst, Monster.com
The session starts with Avinash. His presentation will focus is "Why is '2.0' such a challenge?"
He says you need to alter your mindset, otherwise it is not going to work. The fundamental models of content creation, distribution, and consumption have changed. Content aggregators, user generated content, bloggers, etc. are ways that the model has changed and are what makes it difficult to measure.
He presents us with three ideas to help us deal with this new world.
1. Multiplicity. Think in a multiple ways. Use Analytics, but also feed measurement, Technorati, etc. It's like building a house – you can't use just one tool to build a house. You need to use many different tools to get a grasp of how to build a house, or how to measure what is happening.
2. Unique Measures. You can measure all kinds of things, but are they relevant? For Avinash, the RSS feed is what matters to him. He wants to measure growth in the number of people that have given him permission to push content to him, rather than just number of visitors to his site. The actual day's visits are less important than the growth change over time. You need measurements that are unique to that channel. Think of unique measures, not just old measures that may not be as relevant
3. Unique data collection. Gmail is only one page load, Ajax, videos. How do you measure success? Fake page views have been used, but there are things you can add to the html to measure things. You don't need to pollute the data to measure success.
He sums up by "Get on the train, or get run over."
Jim Sterne is next. He shows how much data is out there and we can measure all kinds of things. We have so much data coming out of our ears.
Web metrics grows up. Evolution went from reporting to benchmarking to analysis to dynamic promotions to hearts and minds.
Search metrics grows up. The evolution of search metrics went from ranking to traffic to analysis to dynamic bidding to predictive buying.
A hitch in measuring this is the economy. There's not as much spending, we're being asked to cut down on costs, which makes it harder to measure things. You need to use the metrics to get better results. I missed some of what he said, my apologies if it's out of context.
Matt Bailey on segmentation. Analytics According to Captain Kirk. Captain Kirk is an analytics genius and pioneer. He talks about green alien women were the key to staying alive on the show.
The huge theme with Matt's presentation is context. Everything needs to be in context. Get away from pages of charts in reporting. You need to instead start with questions, look at complex relationships to make sense of things. Start building context.
Back to Star Trek. Shows stats about total deaths over five years, but no context. Red shirts die more, but need to know more. Keep segmenting for more context, so you can understand what factors are contributing to deaths (aka conversion rates). Look at what you can do to change things. Get alien women! By looking at these segments, we can understand what is increasing or decreasing conversion rate, can make decisions about what to do with this context.
People are not cows. We don't go through a website like a herd. Totally different stories for different people and segments. Conversion rates for one group of people versus another may not be comparable. Again, context is everything. You need to tell a story, it's the only way to compare and contrast what is happening on the site. Get a full time analyst, it makes a huge difference when someone can understand the analytics and make recommendation, it's a huge ROI for the company.
Marshall Sponder. Only has one slide. It's great to not have a scripted presentation, but means that I have missed some more items.. He's taken charge of the social media committee in the Web Analytics Association, now the biggest committee. Started drafting standards, will release later this year. Found search doesn't drive traffic to social networks, but it's things like Digg and Reddit instead.
Traffic to a lot of social networks comes from social media, they don't come from search. One reason is moderators aren't there in social networks to monitor content, add keywords, etc. Makes it hard for search engines to know what page is about.
How to measure what is a conversation?
Social media traffic is more directed than search traffic. Traffic to blogs, especially one he studied in particular, half of a traffic was from social media. That traffic is most directed. Monster and Military.com will release a study tomorrow about how they used social influencers to benefit the Military.com site.
Web Analytics Association has gone out and tried to find out what companies actually need to measure.
You can get people to a site, but you need something for them to do when they get to the site. We can drive people to a site, but need to figure out how to handle them. Analytics can help you with what to measure and what they are doing.
Web analytics in search and social media are similar, but different in one main way. Search is part of marketing, but social media isn't part of anything, not clear where it belongs in the company structure.
Contributor: Keri Morgret