After a long gap, we had web analytics Wednesday again in Bangalore – thanks to Trinadh, Jaisri, Anil and

Ravi. Not to forget our Sponsor – Vijay from Webexpe.

I had volunteered to make one presentation. I started with one topic that I thought would be very useful – “Action oriented web analytics methodology”. The idea was to share how to kick-start web analytics from day one. How to start benefiting from day one. There is certainly a time and place for a proper flow starting with understanding the site objectives to determining the metrics to measuring to optimizing. But I’ve seen many times where the choice of action determines the kind of analytics. For example, I have a writer free for the next 3 days and I got to figure out what to write. Web analytics can come in handy. But, half way through the presentation, I switched tracks and decided to a presentation on Segmentation. The presentation is available here for those who are interested. This is also going to be made available on the wiki on WAW,

Bangalore. I’m still a little embarrassed about the quality of the presentation – both in terms of visuals and content. I had only a few hours to make the presentation.

Lot of people had asked me questions on how to actually start implementing this. I will try to answer their questions through the blog so that it’s available for all. If you have any questions on segmentation, please post a comment and I’ll try my best to answer.

Finally, the process is over. We gave offer letters to 3 brave souls who managed to persevere and finish our tests. One of the critical attributes we were looking at was perseverance. We wanted people who didn’t give up. All three joins us in May 2008 and we will know how effective this test was.

Think about it. What’s common in the name of all web analytics tools? Well, at least most of it. They all are single words formed by joining two words. Check this list

• WebTrends
• ClickTracks
• SiteCatalyst
• IndexTools
• HitBox
• VisitorVille
• WebsideStory
• CoreMetrics
• StatCounter

Notable exceptions are Visual Sciences and Google Analytics. While both of them are formed by two words, they are indeed shown as separate words.  

Interestingly enough, when we chose a name for our tool, we also ended up following this trend. Our tool is called NablerStats.

We are currently working on version 3 and it’s getting better. We hope to make the features public in a few months and start offering it as a solution to customers. 

Can you think of more names that follows this trend? Let me know. It would be interesting to compile the full list.

Last weekend we were at a college and ran our new test with 135 students. It included a standard quantitative test as well as something new that we had created. We have all the students a graph that contained 3 parameters – visits, page views and orders. The graph was for the full month of December and we asked students to come up with their conclusions. I was pleasantly surprised at the quality of observations, considering none of them had heard about web analytics before we met with them.

What was even more interesting is how “the wisdom of crowd” proved true. I have been reading that book for the past few days and didn’t fully agree with the author on all points. But the results of this test was so correct. James Surowiecki would have been happy to see the results. No one student had come up with the best answer. But if I took a combination of all the answers, it would beat the answers of our best analysts.

The short listed students are going to take the test I had mentioned in my previous post. Will keep the results posted here. Meanwhile, if you have ideas / comments regarding this topic, please comment. Would appreciate your thoughts.

The interest in web analytics has been building up and it’s not surprising that one of hot topics on blogs and forums is on “How to hire a junior web analyst”.

We’ve also gone through the same questions – how best do we hire a web analyst. When we are looking for an experienced web analyst, the things to look for are much more obvious – do they understand the web, how good are they with defining metrics, can they interpret and present data, etc. But how do we go about hiring someone who could possibly be a good web analyst one day.  

Here is an interesting idea that we came up with.

Ask the candidate to create a blog, sign up for a free web analytics tool as well as Google Analytics. Then insert the Javascript code in the blog and get to see the numbers. If a candidate who has absolutely no experience is able to reach this stage, it’s a good sign. If a person can’t do at least this much is perhaps not going to make a good fit.

Once the candidate is able to get the code working, the test gets more interesting. We ask them to upload couple of pdf documents, movies, music files, etc on their newly created blog. Then they have to read Google analytics help and try to figure out how to tag these things.

We have just started using this technique. Will post our experience with the test.

Clue for those of you who are interested in joining us – we are certainly going to run you past this test. So be sure that you can do this

P.S. Just in case any one is wondering why there is no web analytics – session 4. We have dropped the idea for now and I’ve take the responsibility of keeping the blog going.