Brian Clifton (PhD) is an independent author, consultant and trainer who specializes in performance optimization using Google Analytics. Brian has been involved in web design and SEO since as far back as 1997, when he built his first website and started defining best practise to advise clients. From 2005-8 he was Head of Web Analytics for Google EMEA, defining the adoption strategy and building a team of pan-European product specialists from scratch.
Recognized internationally as a Google Analytics expert, his latest book, the second edition of Advanced Web Metrics with Google Analytics has just been published. I thought that made a perfect reason to check in with him and catch up on his views of the current state of search analytics.
Brian, congratulations on the publication of the second edition of your book, Advanced Web Metrics with Google Analytics. What has changed in the past two years that made you want to take on the task of writing a second edition—which in fact, seems almost like an entirely new book?
You know, two years ago seems like eons to me in internet years. That’s both exciting and scary, as I feel I am aging in dog years! For example, two years ago social networking meant having a MySpace account and a Facebook page. Twitter was hardly even heard of, browsing the web on your mobile was quite simply so difficult few people bothered and if you used the term MVT most people thought you were referring to music television—not mutli-variate testing!
Now, social marketing is the new SEO. Even the US presidential campaign was won on Facebook (not quite, but social networking was a very large part of fund raising and voter engagement—see the excellent presentation by fellow Xoogler Dan Siroker), eye witness tweets hit the news wire faster than CNN and smartphones are the fastest growing segment of the mobile market.
However, the biggest surprise for the industry was last year’s Adobe purchase of Omniture for a staggeringly ridiculous amount of money in a deal that I still cannot fathom the synergies of. I could go on, but to be honest, even I struggle to keep up to date with our rapidly changing landscape…
Has Google Analytics itself changed that much since you wrote the first edition of the book?
When I started writing the new second edition (last June), my intention was for a an update. A few week’s work—estimated to be ten weeks at most! However, within 3 weeks I was re-writing whole sections and then chapters. This was because once I sat down and gave it some serious thought, so much in the product has changed.
In the past two years, Google Analytics has integrated with AdSense and Feedburner, launched event tracking, advanced segments, Intelligence alerts, motion charts, custom reporting, custom variables, the data export API and the new asynchronous method of tracking. I have probably also missed a few. In the end, the re-write took me 7 months.
Why do you use the word “Advanced” in the title—do I need to understand regression analysis?
As a great friend and mentor once said to me, “Advanced web metrics is about doing the basics very well and applying it in a clever way.” I wish I had thought of that phrase! It epitomizes everything about my approach to web analytics and this book. Thus, I attempted to make this book’s subject matter accessible to a broad spectrum of readers—essentially anyone with a business interest in making their website work better. After all, the concept of measuring success is a universal desire. And no, there is no regression analysis—in fact very little statistical theory at all. To me, web analytics is about measuring the user experience in order to improve it. As long as the reader understands the concept that the larger the sample size, the less likely the data point is a random fluctuation, they will get along just fine.
The reason for “Advanced” in the title was that the other books I have read on Google Analytics have been basic and aimed at the novice user. My book (I hope) takes the user to the next level. That is, understand the setup in detail and how to actually use the resulting data to make a difference to their bottom line.
If you had to choose just one, which Google Analytics metric is your favorite? Why?
Any metric that is monetized is always one that gets my attention—everyone, at all levels of the business, understands the value of money. Hence the Per Visit Value or $Index are great metrics that value a visitor and page respectively. For example, the Per Visit Value metric allows you to home in on what channels are driving your most valuable visitors.
Often I find very valuable visitors are only a trickle in volume while marketers obsess over volume. That is, spending lots of money on campaigns that drive a high visitor count, but have a low monetary value. Clearly such “branding” has its place, but focusing on channels that bring in the highest value visitors provides the greatest benefit to the business. A lot of my book uses Per Visit Values and the $Index metric as the basis for improving website performance.
In your experience, what features/reports are overlooked or underused by search marketers, but are really valuable and useful?
Goals, goals, goals! Defining a goal is what allows you to distill the plethora of report data down to “success metrics.” Measuring success (the title of the first section in the book) is, after all, what web metrics is all about. So, as a website owner/manager, you need to define what success is. This is a goal in Google Analytics terminology as is also often referred to as a key performance indicator (KPI).
Goals are any actions or engagements that build a relationship with your visitors. An obvious goal for an ecommerce site is the completion of a transaction. However, even without ecommerce, your website has goals; for example, the completion of a feedback form, a subscription request, leaving a comment on a blog post, downloading a PDF whitepaper, viewing a special-offers page, or clicking a
mailto: link. Goal conversions are the de facto way to ascertain whether your website is engaging with your visitors. Once set, goals can be monetized, which allows other metrics, such as Per Visit Value and $Index, to be calculated.
Google recently launched the Google Analytics Individual Qualification certification program—an online training course that allows individuals to become certified as a qualified Google Analytics expert. How does the book compare to it?
I am very proud of the GAIQ as it is the most tangible legacy of my work at Google. Back in 2007, while working at Google in London, my view was that there was far too little knowledge “out there” that helped digital marketers, web managers and developers learn the fundamentals of website visitor analysis. I wanted to address this issue using Google Analytics as the example. In my opinion, addressing the education gap for web analytics was going to be the next big step for Google Analytics.
At Google, I had already developed a tiered internal training structure, known as the GAT course (Google Analytics Training), which many Googlers had been through and taken an exam to test their proficiency. However, it was a manual process of workshop education that did not scale well. It was therefore a logical step to produce an automated online version available to everyone. In a nutshell that was the conception of the GAIQ.
Eighteen months and a lot of hard work later, the GAIQ was released to the public in March 2009 as part of Google’s Conversion University program. It was a huge achievement for the European team—the vast majority of product innovation comes from Google’s headquarters in Mountain View. It was therefore good to give back :)
While GAIQ was being developed, I was writing the first edition of the book. What I learnt from building and training the team at Google went into the book and vice versa—that is, what I learnt from writing down my experiences helped me develop the team. So the two compliment each other. The GAIQ focuses on the nuts and bolts of implementation, while in the book that represents three chapters (chapters 6-8). With 500 pages, I have the luxury of covering much more of the analytical reasoning, real-world tasks and hacks that practitioners require when using Google Analytics.
Brian, you have worked as the Europe, Middle East and Africa Head of Google Analytics in the past and now you are an independent author, consultant and trainer. In your opinion, what are the benefits/challenges of working for the giant that Google has become as opposed to being an independent consultant?
I am not sure where this quote comes from, but it sums up the difference nicely:
“At a big company, the process determines your life. At a small company, the people determine your life.”
In my heart I am a practitioner and I love working with good people that can make a difference. You can do that very effectively (make a difference) both within a small company and for a client. That is simply a great deal harder when working within a large organization.
Based in your vast experience during the last decade, what’s next? What should we expect in the next five years from the web analytics field?
Big-picture wise (i.e the web), I would like to see the industry lead the way in protecting online privacy. Unfortunately economics is pulling in the opposite direction. The people that wish to use web analytics data want to know exactly who you are so they can target their campaigns better. You don’t have to be a prophet to know that one day something horrible is going to happen when someone does something nasty by gaining access to your online information—information you probably didn’t even know was being collected. This will stop users sharing their browsing habits—potentially destroying the web analytics industry as we know it today.
The key is putting the user in control of their privacy. The browser developers have a big role to play in this, as does Adobe with its ubiquitous Flash plugin that is so very often used to bypass the privacy settings of a user’s browser. We have a long way to go in this area and I feel governments and legislators will need to be involved. That in itself is scary, as the lack of bureaucracy of the internet has so far been the major factor in the rapid expansion of the web.
In terms of web analytics vendors, some commentators have speculated the next logical step is the merging of web analytics with business intelligence (i.e. the data mining of customer information that already exists in large organizations). However, I am not of that opinion. The exact measurement of customers is contrary to the 95%+ of web visitors that remain anonymous—so you can’t align the data. Unless, that is, the privacy issue becomes “a big brother” issue.
Instead I think we will see more coalescence of off-site web analytics vendors with on-site web analytics providers. That is, combining the big picture view of web activity that providers such as comScore, Hitwise and Radian6 provide, with the detail of what actually happens on your website that vendors such as Google Analytics, WebTrends and Omniture are able to show. In makes sense to do this and some of that has already started to happen. However it is remains disjointed. Expect to see this much more integrated over time.