Keynote speaker Avinash Kaushik, Author, Blogger, Analytics Evangelist tells us how to use amazingly scientific tools to divine the difference between the simply difficult questions of What’s Good and What Sucks on our web sites.
He proposes that there is more things that we can do to with data. It isn’t just stupid log files, that asking questions and finding answers about how people use your web sites is actually fun.
He shows us an early report from Analog. It’s cute! He cites his acronym for measuring “hits” How Idiots Track Success. It doesn’t mean anything…
There’s more data out there than you know what do with it. It’s easy to get a terabyte of data… And most of it’s crap. We should NOT be asking the question “what?”
We should be asking the question “why?”
He shows this great, obligatory slide of embedded circles asking:
1) clickstream – The what
2) Multiple objectives analysis – how much.
3) Experimentation and testing – The way
4) Voices of the Customer &
5) Competitive Intelligence – The What else.
He says the most important thing is that your site is tagged with the google analytics javascript. He recommends that, to check if our site is properly tagged, we should use sitescanga.com
He then talks about “Husband and Wife Data” pairing:
a)how many visits to a piece of content
b) paired with the time spent on that piece of content.
He points out that his will show you which “pieces of content drive outcomes on the site.” Then, you compare this data to the average time spent on the page, and then, in the report, it boils things down to “Good: green. Red: Sucks.”‘
He advocates for us to look at the keyword report… because this shows our visitors’ intent. People are telling us what they want to find on our site… We should be listening!
Next he asks people to consider “bounce rate” for their web site.
We get into an aside talking about bounce rate for people who get to a stream, get what they want, then go. We talk about it for a bit, but then he tells us, what we’re saying is all conjecture, and that we should “let the data tell us.”
He urges us to go home and run this one report in Google analytics to find the “top landing pages” report. Then find out how long people stay on the pages that people go to first….
..the pages that they get into the site from search engines, not from the home page. People do not navigate the way we think they do.
It’s making me a bit uncomfortable. The way he asks us to “find out how much our sites stink” But I have a feeling there’s a reason for it. For us to ask the tough questions and get on with it.
He then asks us to run the keywords report between weeks we find out what the trends are.
He points out that the “Top ten report of anything doesn’t tell you anything. Because top ten doesn’t change.”
Knowing how many visitors come to your web sites is not that useful.
He asks us to think what are the goals of our web sites? For instance, how many people are going to our election page,for example.
He asks us to measure our success by these four analytics:
- visitor loyalty
- visitor recency
- length of visit
- depth of visit
He tells us to ask three questions of visitors when they exit our sites:
1) Why are you here?
You will be astounded! It’s not what you think.
2 ) Were you able to complete your task?
Yes or no. No maybe. If you’re not sure say no.
3) If you were not able to complete your task, why not?
In plain text, people will tell you. Believe me.
This allows us to find segments of discontent.
Once you find completion rates. You will be able to gauge success.
He tells us to “Open a Google ad words account. You don’t have to even buy ads. Then you can use Google Web Optimizer to do AB testing. In six minutes you can find out what works better for your visitors.
He then tells us that to get started with this, all we have to do is “have a one night stand with your IT guy.” Have him or her tag multiple variations of a page, then “watch the green bar move to the right”
He says that if I were to prioritize doing any kind of analytics he would ask us to do this first:
“Talk with your customers.”
He says that the tools are free:
* Google Analytics
* Google Web Optimizers
* Online Surveys
So now, spend your money “finding the right people with the right skills – and pay them,” The tools to do analytics are free. The people to interpret them are what’s expensive.
His web site is called Occam’s Razor
His latest blog post I believe talks about tracking audio. Combining arbitron data and web analytics… intriguing.
His Book is Web Analytics: An Hour a Day
Tags: ima2008
No comments
Comments feed for this article
Trackback link: https://johntynan.com/archives/103/trackback