Search meets Web Analytics at Searchnomics Conference

I was invited as media for the Searchnomics conference, a very important program for the modern web strategy. I’m going to be covering many of the Social Search topics.

Eric Peterson: Web Analytics for Search Marketers “The Hybrid Session”
I’m sitting in the back of the room where the bloggers sit, not because we’re second class citizens, but because that’s where the power is. Avinash Kaushik is here too. Beth, who’s not here, has some info if you want to know more about Eric. I video interviewed Eric, you can check it out on my previous post.

Eric gives a great high level primer for anyone that wants to understand analytics, and runs search marketing:

-Many corporations have employee managed stratgies
-Many web analysts found that the process of web analytics was difficult
-Many web analytics practitioners are considering leaving their job
-Web Analytics is not about technology, you have to manage the software, it’s not people, its’ about repeatable process
-Bounce Rate: How many users leave your sight immediately, couple this info with search terms. “Anyone who looks at your homepage and bounces out is poorly qualified”
-Conversion Rate: Ratio of completed activities to visits or visitors
-Example: Conversion by Search Engine (compare which engine is providing the best results)
-Percent New Visits: Percentage of new visitors, please note that 30% of all cookies are deleted which makes repeat users look new.
-Depth of visit: who’s going deeply into the site, who’s spending a lot of time staying engaged in your site. Also compare by depth of visit by Engine
-Time spent on site: Another proxy for user engagement, but there’s a lot of things that don’t add up correctly. It’s not as a useful as a metric.
-Campaign Cost and Value Metrics: Commonly used cost metrics are: Cost per click (CPC), Cost per acquisition (CPA), Return on Ad Spent (RAOS), Average order Value (AOV)
-Advanced Topics: Visitor Engagement: An index created from: depth of visit and session duration, recently of visit and conversion rate, lifetime visitation history, brand awareness, blog subscriptions and readership, social media interaction.
-Eric publicly teased me for making him disagree with Charlene Li on my video interview, ha.
-Searcher Behavior: Site Navigation
-Closing points: Web analytics and search analytics are not about people, they are about an ongoing process.


Picture 2627Picture 2681Picture 2675Picture 2673Picture 2676Picture 2672Picture 2669Picture 2685Picture 2686

  • Pingback: Web Strategy by Jeremiah » Marissa Mayar, VP of Google on The Future of Search and Announcing Google Gadget Ventures “We pay you $100,000″

  • http://beth.typepad.com Beth Kanter

    I wish I was there, but with your excellent coverage I almost done need to be.

    I have a dumb question. In regards to bounce rate on web sites, you mention that Eric said that “Anyone who looks at your homepage and bounces out is poorly qualified.”

    Does that hold true for blogs?

  • http://www.web-strategist.com/blog/ jeremiah_owyang

    I’m not an expert in this area, but I think Eric was reffering to a website or landing page in a Marketing campaign.

    This probally doesn’t apply so much to blogs, as users come to the index page, and see a lot of content, spend some time, then leave. For blogs, it may be better to look for attention data (duration) and other indicators of engagement (interation)

    I’m no Web Analyst, but that’s my gut reaction.

  • http://blog.instantcognition.com/ Clint

    Jeremiah, you’re halfway there…

    Beth,
    it’s an interesting question although that statement of Eric’s seems to be a generalization.

    That said, a blog home page is somewhat different (generally) than most site home pages in that it carries a lot of the site’s content right there so the drive to get clickthrough from the homepage to other content is less than usual.

    But I’d fall back to the goals of the home page (blog or otherwise). For instance, a click off the homepage that leads to a feed subscription on Bloglines might look like an abandonment (bounce) when in fact the user became more engaged with your content.

    Another difference is that blogs tend to link off to other site/blogs (link love) and there will be a higher concentration of off-blog links on the index than just about anywhere else and if those links don’t spawn a new browser window they will look like a bounce but are they when you are deliberately leading your readers off-blog?

    But let’s say that driving discussion (comments) is a goal of your home page. If the home page has a high bounce rate then it’s probably not doing a great job of starting the conversation.

    Or what if you are supporting your blogging with advertising? If the homepage bounce rate is high, then visitors are not generating a lot of impressions or possibly clicks (unless the bounce rate is driven by adsense clicks).

    Long story short, what are the goals of your blog home page and then what is an acceptable bounce rate within that context?

  • http://beth.typepad.com Beth Kanter

    So, Jeremiah, for blogs we should look at time on site? And thinking outside of google analytics, looking at RSS subs and commenting?

    I’m working on the next screencast for nonprofits – on google adwords and salesforce integration. I’m starting from ground zero – and interview a few folks – what other tips or resources to think more deeply about this? Do you have any relevant blog posts?
    http://beth.typepad.com/beths_blog/2007/06/does-playing-vi.html

    BTW, the google analytics screencast topped over 6,000 views! I have no idea whether that means anything or the tracking programs as there isn’t much standardization in video viewing metrics.

  • http://www.datashaping.com Vincent Granville

    What is the difference between web analytics and web mining? Between text mining and search or query intelligence?

  • Pingback: yet another bounce rate discussion » Instant Cognition » blogs

  • http://www.webanalyticsdemystified.com Eric T. Peterson

    Jeremiah,

    When I was presenting at Searchnomics and talked about bounce rate I was more or less talking about a business web site, one where someone was paying to acquire visitors. Blogs (or blog-only sites) are a somewhat special case where A) few people (correct me if I’m wrong) are paying to drive traffic to them and B) you can get everything you need from a single page view.

    In fact, blogs are such a special case, you can get everything you need from ** zero ** page views, right? I’m pretty well engaged with many of the blogs I read but rarely go to the actual web site, typically only to see what kinds of comments the post has prompted.

    This is yet another reason I push the visitor engagement metric. Visitor engagement is a powerful measurement of qualification that functions independently of simple metrics like bounce rate and time spent on site.

    Nice to see you again and I’m sorry we didn’t have more time to catch up.

    Eric T. Peterson
    Web Analytics Demystified
    http://www.webanalyticsdemystified.com

  • http://blog.instantcognition.com/ Clint

    Eric,
    sorry but even though I suppose it’s theoretically possible to have nothing but a feed, it’s still important to have a site.

    Without a site, folks can’t comment directly on your content (as far as I and, no I don’t think trackbacks count) and although your feed could be subscribed to via an aggregator or WOM, how would you measure the success of such services?

    A blog site, and it’s homepage offer the easiest way to understand your content’s ability to convert subscribers. It offers your users ways to interact that just isn’t available through other methods.

  • http://www.shinylittlegadgets.com Heath Weaver

    Eric answered the question above already, but maybe I could add that our bounce rate metric is looking at people who stayed shorter than 5 to 10 seconds. No matter what the site (blog or not) if someone comes off a search and leaves that quickly then they didn’t get what they hoped for or expected. IMO.

  • http://blog.instantcognition.com/ Clint

    Jeremiah, you're halfway there…

    Beth,
    it's an interesting question although that statement of Eric's seems to be a generalization.

    That said, a blog home page is somewhat different (generally) than most site home pages in that it carries a lot of the site's content right there so the drive to get clickthrough from the homepage to other content is less than usual.

    But I'd fall back to the goals of the home page (blog or otherwise). For instance, a click off the homepage that leads to a feed subscription on Bloglines might look like an abandonment (bounce) when in fact the user became more engaged with your content.

    Another difference is that blogs tend to link off to other site/blogs (link love) and there will be a higher concentration of off-blog links on the index than just about anywhere else and if those links don't spawn a new browser window they will look like a bounce but are they when you are deliberately leading your readers off-blog?

    But let's say that driving discussion (comments) is a goal of your home page. If the home page has a high bounce rate then it's probably not doing a great job of starting the conversation.

    Or what if you are supporting your blogging with advertising? If the homepage bounce rate is high, then visitors are not generating a lot of impressions or possibly clicks (unless the bounce rate is driven by adsense clicks).

    Long story short, what are the goals of your blog home page and then what is an acceptable bounce rate within that context?

  • http://www.monitorlcd17.org Monitor LCD 17

    I feel you heath