A conversation with Avinash Kaushik, Web Analytics Guru
Avinash Kaushik is a leading expert in the new field of Web Analytics. His blog, Occam's Razor, is one of the most popular blogs on this subject. He has lots of other exciting things happening: he's the Analytics Evangelist for Google, author of the book Web Analytics: An Hour A Day published by Wiley, and most recently, is a co-founder of a startup, Market Motive, focused on spreading knowledge for Internet Marketing. He was kind enough to agree to an email interview, given below.
If you are interested or involved in Web Analytics, I guarantee that his answers will give you much to think about.
Q - How did you get into Web Analytics? What is it about this field that attracted your interest?
AK - I ended up in Web Analytics by pure chance.
My former roles were in decision support systems, both on the business and technical side of the fence. The Intuit job, my foray into web analytics, was attractive more because of the people and the company.
But I had always been fascinated by the web and the job allowed me to put my experience in decision support with the fantastic piece of art that the web is.
At some level it was lucky to get into web analytics with no baggage or hang ups or having read any books, it allowed me to bring a fresh and completely different perspective to it.
Q - In your study of web site user behavior, what are some of the most surprising results you've found?
AK - I am surprised that even in 2007 given how pervasive the web is and how it is used that we continue to obsess on conversion rates, essentially solving for a minority of site traffic as if people came to our sites for just one purpose. That is so 1997.
I am frequently humbled by the lessons customers have taught me when we listen to them using surveys or multivariate tests or site visits. Cool and sexy is not always enough. Simplicity is the key. Solving for customers and bottom-line is possible. Having clear calls to action on all pages (especially on those where there is no "add to cart button") and the importance of solving for your customer personas (just look at www.newegg.com, no one will call it the prettiest site in the world and yet it consistently outranks www.apple.com and www.amazon.com when it comes to customer satisfaction!) cannot be emphasized enough.
Q - With the benefit of your deep background in this field, what do you think of the Google Analytics product: What are its strengths? Which types of companies is it most useful for? Which areas do you think need to be improved?
AK - I have just published a comprehensive review of all web analytics
vendors - link:
http://www.kaushik.net/avinash
In the video I mention two key strengths of Google Analytics:
1) Data Democracy: Google Analytics is a drop dead easy tool to use and presents a lot of complex web analytics data in a very easy to understand manner. Because of this it flips the traditional web analytics model were a few people in the company had access to the data and shared it with others. With GA you can give everyone access to the tool and they can help themselves.
2) Best of breed search analytics: The reports and segmentation options you'll find in Google Analytics to analyze your site's search data is really good. Perhaps it should not be surprising that a web analytics tool from a search engine is good at that. You don't have to tag your campaigns because of auto tagging which saves hassle and improves data quality. Your data is also imported and integrated and presented with some unique reports.
In terms of who GA is right for...... Google Analytics is right for any company that will benefit from the above two features. The nice thing is that unlike the past were you can rule tools in and out on paper, now you don't have to take a random person's, or a "guru's" opinion, on benefits of the tool. GA is free. Throw it on your site and try it for yourself and using your own data from your own site you can determine if it is right for you.
In terms of what needs improvement.... Currently GA provides 27 pre-built segments that you can apply to any of the 80 odd reports to get 27 times 80 sets of segmented data. But I am selfish. I would love to have even more flexibility when it comes to creating visitor segments that are most relevant to each business.
Q - Your blog, Occam's Razor is one of the most successful blogs in this field. What has blogging meant to you? Are there things you would do differently with the blog if you had to start over?
AK - My wife's opinion is that the blog is our third child. :)
When I started writing the blog a little over a year ago my hope was to have around a thousand visitors a month because that is how many people I thought were my core target audience. Yesterday the number of RSS feed subscribers were at 4,600 and there were 30,000 visitors last month. That in many ways simply astounds me.
These numbers also mean that I feel a deep sense of obligation to the people who read the blog. There is always a pressure to deliver the highest possible quality in the posts that my humble skills will allow.
The blog means the world to me because of the conversation that I can have with people from around the world (around 30% of the site traffic is international). All these wonderful people write comments and their own perspectives which I learn from and all these comments add to the conversation (on my blog visitors have written approximately as much content as I have written, word for word).
In terms of different..... I wrote a post at the end of the first month I think, I would not have written that in hindsight. But other than that I would not do anything different, the blog has managed to stay hyper focused on what my initial vision was and I think it works well.
Q - Even now, Web Analytics is seen as an afterthought in some web companies, rather than being an integral part of the business process. How do you convince these companies of its importance?
AK - I agree with you, it still exits in silos (both from organization and data perspectives).
At some level it really requires the business realizing the importance from the inside that matters. No amount of outsiders coming and pontificating can drive fundamental change.
If you are inside the company then you have an inside track to helping your company realize the value of web analytics. My advice would be to focus on two simple things in a very hard core way: 1) value the web can deliver to the bottom line and 2) value the web can deliver to your customers. The interesting thing is that the web can do both of those in an efficient and scalable manner, unlike any other channel.
And if you want to help your company do #1 and #2 you need web analytics. Start showing it in small ways (rather than trying to create a overnight revolution, those rarely succeed) and I assure you that your company will "get it". Few people can argue with profit and fewer still can argue making customers happy.
Q - You've just launched a new company, Market Motive. Can you tell us more about it? Who are your target customers? Will you be offering any free content, or is it all behind the "paid" curtain?
AK - Market Motive's mission is to focus on helping Online Professionals be massively successful through access to the latest best practices and insights from the best people in each discipline. We hope to deliver that by providing fresh and unique content that will only be available at www.marketmotive.com.
The initial areas of attention will be: SEO, PPC/SEM, Web Analytics, Conversion, Email Marketing, Online PR, and Marketing Processes. We will provide videos, podcasts that provide a unique way to learn, these will be complemented with live phone-in sessions were subscribers will be able to ask their questions and get them answered by the dream team.
The target audience is Professionals whose job it is to deliver for their companies in any / all of the above mentioned seven areas.
The content created at Market Motive will be available on an unlimited consumption basis to only the subscribers. All the faculty have blogs on which they are very active.
Q - What advice would you give to a small company that's just starting to get deeper into Web Analytics (beyond basic Page Views and Referrer URLs)?
AK - Use your web analytics tool to answer questions and not simply measure "KPI's".
Here are the three questions to answer:
1) Where are people coming from? (Referring URL's, Search Engines, Key Words, etc) This helps you infer intent and identify valuable sources of qualified traffic (by simply measuring bounce rate).
2) What do they do when they are on my site? (Content Consumption, Top Entry Pages, Top Visited Pages, Site Overlay etc) This helps you understand what people might be looking for and is it easy to find and is it what you want them to see.
3) What were the outcomes, both for you and the visitors? (Revenue, Conversion Rate, Task Completion Rates, # of leads, Likelihood to Recommend, Customer Satisfaction etc) Your site should make a difference to their existence. Is it?
Q - What is the biggest mistake in the use of WA? What should people watch out for?
AK - Usually the weakest link is that website owners rarely sit down and define why their site exists and if that's the case then any metric will look like success. You should be able to answer in fifteen words of less "why does my site exist" and then be able to identify two metrics that help you measure if your website is delivering.
The other big mistake is that Marketers and Website Owners think that they represent their customers. This is mostly false. We, company employees, are too close to our companies to ever be able to think like our customers. If you want to know what your customers think of your website experience, ask them.
Q - What radical changes do you think we will see in Web Analytics in the next 3-5 years? Do you expect to see a big impact from the proliferation of Social Networks (like Facebook)? What about SEO and the increasing importance of search engine traffic?
AK - The web reinvents itself and that is what makes it fun. I think with all the web 2.0 buzz we are in the middle of one such transformative experience. Each such transformation like that requires the measurement methodologies to evolve as well. We are now trying to figure out how to measure ajax and flash and videos and podcasts and so on.
In the next couple of years I think web analytics will change radically. In the near term we will evolve to measure the aforementioned fluid experiences much more effectively. In the slightly longer term I am anticipating (and hoping) that web analytics will transform into business analytics. A way of life, a normal way of existence, just like other pieces of analytics that tend to have nothing special about them, and not an afterthought.
I have recently written about Web Analytics 2.0
(http://www.kaushik.net/avinash
As regards to social networks and SEO etc I think that these types of wonderful things will never leave us (hopefully not). From a web analytics perspective we need to come up with more efficient ways to collect data, not matter which way life on the web evolves. I am optimistic that in the next few years we'll have that figured out.
Previous related articles:
Top 8 Reasons to Implement Tracking and Measurement for your Web Site
Web Design and the Scientific Method
A conversation with Guy Kawasaki
