Interview with Analytics Evangelist Avinash Kaushik (part 1 of 2)

For five years Compete has invited marketing pioneers from across the US to participate in its annual Client Forum. Among this year’s presenters was author, blogger and analytics evangelist Avinash Kaushik. He entertained and educated us on how to turn web analytics into actionable insights. I decided to interview him for the blog to share what we learned with anyone that could not attend.

During the Compete Client Forum you spoke about the flaws in just using traditional metrics for measuring online marketing and the critical need for marketing strategies to deliver actionable insights. What are some of the pitfalls that online marketers fall into?

The biggest one is that unfortunately a lot of online marketing remains a faith based initiative.

In one of the most data rich environments in the world, the web, faith based initiatives are a sub optimal outcome.

The flaw is less that Marketers are using traditional metrics, it’s more that the world view of what should be used and what might provide the most amounts of insights has been severely limited.

The web is astoundingly measurable, and while it is not perfect, the insights that you can get from your web data far surpass those you can get from any other channel. At a very low cost to boot!

Here’s an example: Just compare the measurement of success of a GM Chevy advertisement in Time magazine and on Chevy.com (or a banner running on yahoo.com). In the former case you know how many subscribers Time has and you also know how many people buy it. At best you have a measure of possible "impressions". At best. Now in case of the latter you know exactly how many people saw it, how many interacted with it, how many go to the site and maybe configure a car, and… a lot.

Online Marketers have just started to understand the power of data on the web, the onus is on all of us to lean, make things easy, and show much more of the web is accountable.

What is Web Analytics 2.0?

Traditional Web Analytics was all about clickstream data. Log files. JavaScript Tags. HITS, Visits and other such nice stuff. It helped us progress to understanding What was happening on our websites.

But even though we all did our best it did not yield the kinds of insights that would be actionable and, dare I say, earth shattering.

The problem was that all that data was sub optimal at explaining Why. And that right there was the key.

Without the Why, the What is not very actionable.

So a little while back I proposed Web Analytics 2.0, the next generation approach towards getting both the What and Why answered. Here is the official (well my) definition:

Web Analytics 2.0 is:

  1. The analysis of qualitative and quantitative data from your website and the competition
  2. To drive a continual improvement of the online experience that your customers, and potential customers have
  3. Which translates into your desired outcomes (online and offline)

More at webanalytics20.com.

Everyone is making promises around their Web analytics solutions — what’s the bare minimum that should be included?

Here is the great thing, 80% of the data you will get out of any web analytics tool is the same. So it would be really hard for you to screw up.

The bare minimum that is included probably should be the 80 or so standard reports that you’ll get out of the free Google Analytics solution or the free Yahoo! IndexTools solution or the one from Microsoft. That means reports that show you traffic sources and keywords, allows you to track your paid search campaigns as well as organic traffic, easily illustrate what content is consumed and is of value (and what content is yucky and bounces traffic!), the ability for you to segment your data (at least to some extent), empower you to track rich media (flash, ajax, video etc) without having to create fake page views and of course finally a full complement of ecommerce tracking.

You can get all that for free now, so when you buy a solution it is a good idea to request that you get more than everything above.

And you can. For example some of the paid solutions provide a great ability for you to bring complex sets of data into the tool (beyond clickstream), others thrive on their ability to bundle testing solutions, others still provide you the ability to truly do some amazing and advanced segmentation.

Identify your needs, make sure you are not paying for features and data you get for free, then buy the tool that is right for you.

Here is a post you might find relevant in that process: Web Analytics Tool Selection: Three Questions to ask Yourself.

What are some of the free analytics tools that you use and like?

I have eight tools on my blog. I guess I am "special". :)

I use Google Analytics and IndexTools amongst the free more widely used tools. GA is simply the most easy to use tool and I am simply loving how people are creating plugins on top of it (like the one report I really wanted "What’s Changed"). IndexTools is so great to create custom reports, when I have a fast question I just throw the Metrics and Dimensions on a "page" and boom (!) I have what I want, ten seconds!

I also dabble in ClickTale and CrazyEgg. They both have free versions and while they don’t have the depth of standard WA tools, each answers an interesting question uniquely.

While it is not free I am rather fond of ClickTracks. It is not very expensive. My fondness is because for such a low price I can get infinite segmentation capabilities (some call it "n dimensional" segmentation) and also the ability to "reprocess" data.
Now I don’t really need to do that on my blog, after all it only gets 45k Visitors a month, but it is fun. :)

Stay tuned for more Q&A with Avinash"¦