Archive for October, 2009

Oct 22
2009

Google has just released version 6.602 of their Urchin Software product. Version 6.602 contains one major new feature, along with several small improvements and bug fixes. The major new feature in this version is multi-language capability. Now our friends from Germany, Spain, Mexico, Japan, and other regions can have localized instances of Urchin. Very cool :)

Here is a quick overview of the features included in this release:

  • Switch between supported languages (total of 10)
  • Added SSL/TLS protocol support to External Authentication (LDAP)
  • The Urchin “Home” page (sometimes called the Roll-up Report) now displays metrics for all profiles visible to the logged in user. Data can be shown in 3 views: Basic, Business, and Admin.
  • Easier one-click installation: New one-click installation which also sets up the included postgresql database server at the same time! This reduces much of the setup time, since the database server no longer has to be setup and configured. The user can still opt to use existing database servers if desired.
  • Optimized GeoDB. You can now select between a full GeoDB or a lighweight GeoDB providing metro information up to the country level only. This significantly reduces the memory footprint.
  • The zip utility has been updated to work with data files > 2 Gb.
  • New UI tweaks have been made on the Marketing Summary report so that the user can sort metrics sources, keywords, and campaigns by their appropriate metrics.
  • Remote log file downloading. A new configurable parameter for remote log download timeout has been added to urchin.conf.

In addition, many bug fixes in the areas of log processing, licensing, migration, and embedded help were included in this release.

Important notes:

  • Urchin 6.6xx no longer supports the MySQL 4.x.x. database. If you are using this version of MySQL, you’ll need to upgrade to 5.03+.
  • FreeBSD 5.x platform has been deprecated.

That’s a lot of new stuff for a point release upgrade :)   This new version has currently only been released to Urchin Resellers.  If you would like a trial copy, please contact us, or you can download version 6.6.0.1 (the current version) here.

Oct 20
2009

Marketers and analysts, are you ready for the latest version of Google Analytics!!! :)

As announced today on the Google Analytics blog and at eMetrics, the Google Analytics team is rolling out a new version of GA (version 4) with exciting new capabilities, many more customization options, and the totally awesome and super powerful Intelligence feature.

Over the next several days I’ll be posting specific posts about each of the new GA features, but for now here is a quick summary of what you’ll start seeing in your reports.

  • Expanded goals
  • Mobile reporting
  • Advanced table filtering
  • Unique visitor segmentation
  • Multiple custom variables
  • Sharing of custom segments and custom report templates
  • Intelligence
  • Automated and custom alerts

Please note that some of these features are rolled out as we speak (ok, it is past noon Pacific, 10/20/09 :) ), and others will be rolled out in the next few days or weeks.

More Power!

  • Goals and more goals! We have conveyed your requests to Google. You weren’t satisfied with 4 goals and now you can have up to 20; yes, 20 goals! Measure those important actions! In addition, you can now measure two engagement metrics, Time on Site and Pages per Visit.
  • Expanded mobile reporting. Not only you can track traffic from your iPhones, but now GA allows you to track mobile applications built for iPhone and Android devices (great way for application developers to measure engagement). In addition, traffic to mobile sites can be tracked (for sites built in PHP, Perl, JSP, ASPX), for devices with or without JavaScript.
  • Secondary dimensions and pivoting. Great table and filtering features. Check out our post on these here.
  • Unique Visitors Segmentation: now available across many dimensions in your Custom Reports. Answering the question of “how many actual visitors” got a little bit easier.

More Flexibility and Customization Options

  • Multiple custom variables. The requests of the analytics ninjas have been answered! Define an important interaction for your site and set a custom variable to track it. Tracking can be done at the page, session, and visitor level.
  • Sharing custom segments and custom report templates. If you can create them, you can now share them :) . It’s as easy as copying and pasting a URL.

Intelligence

  • Analytics Intelligence. This new feature is just great! Any time a tool can help automate the mundane and allows you to focus on the business, you gotta love it! This new report will provide automatic alerts when significant changes in your data occur, in daily, weekly and monthly intervals.
  • Custom alerts. You can create your own custom alerts and be notified by email or through the UI.

That is it for now. Log in to your GA account and check out some of these features and let us know what you think!

Oct 18
2009

It’s one day away, eMetrics will be in Washington DC, and it’s not too late to register (you never know, check with your boss for some unused training budget :) ).

Jim Sterne and the eMetrics team have an incredible line up of speakers and tracks, including multi-channel metrics, search analytics, emergent media and many others.

In addition, there is a number of workshops offered. Topics include: reports, predictive analytics as well as a workshop on Google Analytics by our dear friend and GA Ninja Caleb Whitemore of AnalyticsPros.

So don’t miss out if you can make it, you can still register here.

Oct 16
2009

Third day at the Google Analytics conference at Google. The day started off with a talk by Hal Varian, Chief Economist at Google. Hal is just at another level when it comes to analysis! He does a very good job at taking something complex and make it simple for the rest of us to understand :) .

Didn’t take a lot of notes (I was mesmerized most of the time) but here are some takeaways:

  • Data is available and relatively cheap but we still need humans to analyze and make sense of all these piles of data
  • Check out this graph about the US recession, http://www.slate.com/id/2216238/, and I’m sure you’ll agree it is not a pretty picture!
  • I really liked his comment, “if you torture the data long enough it’ll confess to anything.”
  • If you haven’t already, check out this very cool and super useful tool, Google Search Insight. With this tool, one can assess patterns of search volume. you can select a specific region or categories, time frames, and properties. See seasonality impacts on keywords related to what you sell, compare search volumes, and more. Don’t forget to examine the “forecast data”.

Oct 14
2009

Second day at the Google Analytics Summit here at Google and I am surrounded by some of the brightest and most intelligent gents and ladies in the industry. Avinash Kaushik, Google Analytics Evangelist, is the keynote. The topic is titled “crime against humanity”! :)

With his usual charisma and humor, Avinash covered a number of areas including the process of making decisions and smart analytics. I’ll summarize a few points in this post. This is not meant to be comprehensive so to get more, buy Avinash’s new book! :) ) .

  • Websites are not the center of the universe anymore and neither is centralized analytics tagging. The model used to be Content Creation  —>  Content Distribution  —>  Content Consumption. This model is changing. The way we create content, the way we consume it is changing (blogs, tweets, etc.); marketers and analysts don’t think about this issue enough. Just looking at your site analytics is not going to be sufficient. The data that we need to make marketing decision effectively is more distributed and not in one centralized place.
  • The challenge is going to be to figure out how to integrate all this data from all different systems automagically :) as Avinash put it; pull the data in, scrape, integrate, and then correlate and compute.
  • Vendors, practitioners, and consultants should start thinking beyond existing (read this as old) metrics: (visits, pageviews, time on site, etc.), and look into things like engagement, reach, velocity, demand, network strength, and activity. As we have moved away for hits (hits = How Idiots Track Success :) ), we need to move to smart metrics.
  • An example of smart analytics for measuring twitter. Move beyond just number of followers, look how my message that I am communicating is amplified. Look for number of retweets per thousand followers. Twitter is not an ad – it’s about conversation and not shouting. Conversion rate is another cool metric. Numbers of replies sent/day and replies received/day indicate you are communicating with people.
  • Another example for measuring a success of a blog. Go beyond pageviews! RSS/feed subscribers is much more important than visits or visitors. Metrics such as conversion rate per post, words in posts, words in comments all are very insightful.
  • And finally think of outcomes and behavior that adds value to business.

Oct 13
2009

I’m at the Googleplex this week participating in a summit. One of the keynotes is Dan Siroker, Former Director of Analytics for the Obama Presidential Campaign and is currently co-founder of CarrotSticks, a website aiming to improve math skills in kids.

Yes the election is over and this might sound like an out of the date post, but actually the message here is that Dan went on to found a start-up with the knowledge and expertise he gained from the Obama campaign. What he accomplished and learned in the political campaign is very applicable to the private sector and hopefully we can all learn a thing or two.

Some interesting facts about Obama is he out-performed McCain in the “new media” category significantly, in terms of Facebook friends, YouTube views, and website unique visitors, as well as the very important aspect of fund raising. The money raised for Obama was $656 million versus $201 millionfor McCain (not counting the Federal funds). Out of the $656 million, a staggering $500 million was raised online.

Analytics Lessons Learned

  1. Define success and also define quantifiable success measures.Metrics, like cost per click, email sign up rate, and revenue per email recipient. In Dan’s startup, he is now measuring: cost per click, cost per playing user (free account), revenue per paying user (paid account). So when the revenue per paying user exceeds the cost per playing user, they make money! :)
  2. Question assumptions.Testing is king, A/B and multivariate testing both play a key role in deciding what call to action, what creative, and other aspects work best!
  3. Divide and conquer.Segment users, for example, those that never signed up, those signed up but never donated, and those that previously donated.
  4. Don’t reinvent the wheel.Use available and free tools like Google Apps, Google Websize Optimizer, and Google Analytics.
  5. Take advanatage of circumstances.Whether it is internal to your business or something external (economy, competitiion, etc), give your visitors current and relevant content.