Q&A With Avinash Kaushik, Google Analytics Evangelist & Customer Insight Guru
Avinash Kaushik is Google's well-known and widely respected analytics guru. In this wide-ranging interview, he talks about his passion for metrics, why they're critical for success, and how search marketers can use analytics to take their campaigns to the next level.
You’re the analytics evangelist for Google. Doesn’t every marketer understand the importance of analytics? With hundreds of thousands or perhaps millions using Google Analytics, why would Google need an evangelist, and what do you see as the most important part of your job?
Numbers are hard to come by on this but in my humble experience a tiny fraction of people who should use data productively access it, and a tiny fraction of that actually end up using data effectively. We, as a universe, have a long way to go.
My role at Google is in two parts. In the inward facing part I am the “customer evangelist” as I help shape the vision, direction and features of 13 different Google tools that provide data to customers. In the outward facing role I help the top xx Google customers to leverage data more effectively.
The most important part of my role is that I am a small part of larger effort to create a data democracy in the online world.
All of the above is distinct from my role as a blogger (evangelizing the use of data in web decision making) and as the co-founder of MarketMotive (providing latest online marketing education and certification).
Avinash, following your first book, Web Analytics: An Hour a Day, what drove you to write your second book, Web Analytics 2.0: The Art of Online Accountability and Science of Customer Centricity? Did the field change so much since the first book?
The world indeed has changed a lot in two years, especially in three areas: user centric design, competitive intelligence and social media. There are a raft of new and delicious options that simply did not exist when I wrote the first book.
But the primary impetus behind writing the book was to address challenges that we all now face, challenges that present new opportunities (to engage and influence current and future customers) and how to measure success is a complex ever evolving ecosystem.
An example of that last point is in Chapter 9, if you see Figure 9.04. It makes you stand back and marvel at how we are measuring anything at all!
Even in a “standard” area like paid search analytics there has been so much evolution in the last couple years, analytics of which are covered in the new book.
Do you see both books as an evolution? I mean, would you recommend readers start with the first and then go to the newest one?
It is definitely an evolution.
People who have read the first book should feel that the second book is an immediate step up to a more evolved way of thinking about analytics, from the classic Trinity to the new Web Analytics 2.0. With each chapter there is a new way of thinking about what we already know and assumed. The second half of the book is where all the delicious stuff is that will help you change the game—it covers analytical techniques, social media analysis, competitive intelligence, new ucd approaches and of course things like multi-touch campaign attribution analysis.
If you want to start with the basic and take a gradual course then I recommend Web Analytics: An Hour A Day, but if you are willing to be a bit brave then Web Analytics 2.0 will get you to the goal faster.
I did want to point out that both the book are written for marketers, executives and analysts. You don’t need particular deep technical knowledge to become a analysis ninja.
Can you describe in a few words what is the main philosophy behind the book, the concept of Web Analytics 2.0?
Here is my definition of Web Analytics 2.0.
It is: the analysis of qualitative and quantitative data from your website and the competition, to drive a continual improvement of the online experience of your customers and prospects, which translates into your desired outcomes (online and offline).
Put simply, it is the art and science behind making intelligent decisions about all you do online—as a company, blogger, non-profit.
One sentence in particular caught my attention: “We’ve evolved from hits to page views to visits. Now we have outcomes.” Can you elaborate on that?
I have become convinced that more of the online world is not data driven because we have been reporting “silly metrics.” By that I mean hits and visits and page views and even visitors. What is the point of all those metrics?
They are all “aggregate” metrics that simply tell you “consumption”.
What they don’t do is answer this question: “What the hell happened?”
That question is important to answer because that is what matters to every senior decision maker, people who cut checks, people who will promote you and me.
Hence my near paranoid focus on first measuring outcomes—what happened as a result of all those people showing up? Did we make money? Are more people coming to my protest against cutting redwood trees tomorrow? Am I wasting my time marketing on twitter? What was the point of using video to sell micro-chips to B2B customers of Texas Instruments? So on and so forth.
God will welcome into sweet heaven people who focus on measuring outcomes. Because outcomes help businesses and people get better each day.
Your book has a lot of great advice for big companies trying to understand how their customers interact with their websites. But if I work for a Small-Medium Business, with a very limited budget, how can I go about implement this strategy? Any advice on whether search marketers should hire skilled web analysts or outsource the job to experts?
The book should be useful to businesses of any size. Throughout the book there are recommendations where to start and what to do first or what tools to use. For example on Page 13 it tells you that if you are a small biz then you must do x first, then y and then z and don’t worry about a and b. Or in Chapter 10 The Ladder to Analytics Nirvana gives a very specific road map for someone who is small to someone who is big. Same thing with Paid Search, Chapter 4 has the basic to medium stuff in terms of what to analyze for higher ROI and Chapter 11 is where all the juicy complex “I am going to be awesome” advanced stuff is.
My hope was to always provide a shallow end of the pool so everyone can get in, then those who want to do more can slowly, with confidence, move to the deeper part.
In terms of hiring… it would depend on your budgets and in-house sophistication. I have come to believe that if you don’t know what you are doing it is best to hire a consultant and put them on a “profitability plan” (i.e. you do the work I don’t know and we’ll both share the profit—not just hourly rates). Over time as your budgets increase, you become a medium sized biz, it is prudent to bring it all in.
Another interesting subject you deal with in the book is the web analytics career. What do I need to succeed as a web analyst? And if my main focus is on SEM, how can your book help me succeed?
You need to loooooove the web and all the glory and all the possibilities. If you don’t have passion for this medium there is no way you can put up with the work that is required.
Other main skills I look for: Initiative. Curiosity. An aptitude for data. Statistics 101. Self taught.
If your focus is on search then first Chapter 13 will help you plan your career effectively and help you create your own path for success. But most of all I am sure at some level we all understand that Search is not everything, certainly not paid search. The book will help you understand how the broad portfolio of online marketing works and of course how to be king by being data driven.
In the end it should make you a more rounded individual, and thus, I hope, a more marketable person in the job market.
You’re well known for the 10/90 rule. Why do you think it’s more important to put the emphasis on people rather than technology? How much is enough? How do you set reasonable goals and know whether you’ve achieved them, or to put your head down and try, try again?
Here is the picture that illustrates my concept of Multiplicity:
What it shows is the breadth and depth of the tools that are required to answer the four important questions: What, How Much, Why and What Else.
Even a few years ago for you to get access to tools would have required you to spend a lot of money, that’s not the case any more. Clickstream? Surveys? Competitive intelligence? A/B or multivariate testing? You got it, every single one has a 95% world class tool available for free.
So having tools (access to data) is no longer the key differentiator between companies, large or small. Having the brains to actually make sense of it all, look to the right tool to get the right answer, be able to actually analyze the data and not just data puke is not cheap. That’s where humans come in, that’s where the strategic differentiator comes in.
I had created the 10/90 rule almost five years ago when I was at Intuit. Never in my wildest dreams did I think it would actually be practical, but it is now. Every company in the world should not shoot for 10/90 (10 tools and 90 in people), experiment and find your balance. I think many people start with 10/90 and in a few years might morph to 35/65. No worries as long as you can so ROI impact.
But I have to admit, if you are not egregiously overloaded in the big brains (internal hires or consultants) you don’t stand a chance.
There’s a lot of buzz around attribution modeling, knowing the “value” of certain clicks to the overall conversion process. Is it important to know whether the last click was the major influence on a conversion, or will analytics packages increasingly try to understand the various steps in the overall buying funnel? If so, how?
This is the only question on which I’ll bow out from answering, it is complex and I think the reason we are in the soup we are is we look for shortcut quick answers. There are none. Of course I absolutely apply critical thinking to this in the book and provide answers.
Analytics systems are notorious for delivering vastly different interpretations of seemingly simple data—how users interact with a site. Why aren’t there more standards, and why are the reports from different vendors so different?
This is like asking a four year old boy how come he is so lame that he does not already exhibit the mannerisms and sophistication of a fully grown man.
Our industry is a baby, it is in a growth spurt, we must be patient and let things evolve. And they will.
Complete side note: It is utterly futile to wait for perfect data to make decisions and / or spend time comparing numbers between Omniture and WebTrends. What is the point of it? So we are more comfortable that one piece of data is 5% better than the other? Pause and think for 60 seconds how tv ratings are measured. It will horrify you how the data is collected and subsequently used for multi million dollar decisions. On its worst day the worst third party cookie based tool gives better and more accountable data for Marketing spend online. I personally don’t recommend wasting time trying to get the last 5% accuracy, simply not worth it. Implement tools correctly and completely. Don’t worry about the wife you just divorced. Worry about the one you just married and make a happy life with her.
Thanks so much for the opportunity to do this interview.
Thank you for spending the time to map your most interesting views of this subject, Avinash.
Editor’s note: Want to know more about Avinash’s new book, Web Analytics 2.0? He’s written a short overview of the book in this blog post. You can also buy the book through this affiliate link with Amazon and Avinash will donate 100% of the proceeds to two charities, the The Smile Train and Ekal Vidyalaya.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.