Web analytics books and resources

Whitepapers by Brian Clifton

This 19 page vendor agnostic document details every possible accuracy consideration for on-site web analytics tools such as Google Analytics, Omniture, Webtrends etc., and how to mitigate these and get comfortable with your data.
>> Undersanding Web Analytics Accuracy



Tracking offline marketing whitepaper When it comes to tracking offline marketing campaigns, many marketers are unaware of the potential of using their existing web analytics tool to measure success. In this PDF whitepaper I describe how to use four techniques with Google Analytics ? two technical (redirection) and two non-technical ? for measuring the success (or not) of your offline marketing campaigns.
>> Tracking Offline Marketing Campaigns with Google Analytics


Highly recommended reading These are all books that I have read and over the years. They have helped me to learn and develop skills for understanding the web, its measurement and web site analysis. They have influenced my thoughts and solidified my own personal vision. Links open a new window to Amazon.


Thinking, Fast and Slow by Daniel Kahneman


Easily the best work book I have read…

There are plenty of analytical books that can teach you new facts – new ways to collect data points, say something about it, plot it and visualise it. However, very few books get you to rethink how you process raw information. This book does. It made me rethink how I look at data, what biases are in-built to my thinking and how to recognise the common error of inferring causality when only correlation exists. This is not a light read, and I disagreed with some of Daniel’s points (can you disagree with a Nobel laureate?)*. However, any book that is capable of getting a seasoned analyst to reassess their thinking is worth its weight in gold as a piece of educational literature.

Example excerpt: “We can’t assume they will learn anything form mere statistics. Lets show them one or two representative cases to influence their [thinking]”

Before making this statement at the end of Chapter 16, Daniel describes the evidence from psychological experiments that data on its own is useless – you need to tell a story in order to influence people’s thinking i.e. get them to make a change based on the data. Great to know there are studies to support what intuitively analysts already know.

*What I disagreed with
Chapter 11 “The Anchoring Index”. The calculation of this appears odd to me i.e the ratio of the two differences. Surely, the difference from what people estimate without any anchor present in the text needs to be taken into account?

Most memorable quote
“Marital stability is well predicted by the formula: Frequency of lovemaking minus frequency of quarrels.”

You don’t want your result to be a negative number!

signal-and-the-noiseSignal and the Noise by Nate Silver
amazon.comThis is a similar book to Fooled by Randomness (from Nassim Taleb), that is, a very readable book about chance and risk and how humans are poor predictors of the future. I revisited the subject some 5 years after Nassim’s book based on the reputation of the author – Nate shot to fame by predicting (very accurately), the state by state results of two US elections by applying Bayesian reasoning to polling data. This is a “punchy” book in the same way Freakonomics was. Bayesian reasoning is at the heart of of it. I particularly enjoyed the oft-repeated references to the dangers of big data being the cure for all i.e. just because there is lots of data, it doesn?t mean we can make good predictions. In my experience, most big data is actually poor quality or irrelevant data i.e. noise. Hence, the false sense of security big data provides is actually dangerous.
Web Analytics 2.0 by Avinash Kaushik
amazon.comI have known Avinash since 2006 and he has honoured me by writing the forward to two of my books – the second edition of Advanced Web Metrics and Successful Analytics. The guy is an inspiration to hear speak and never ceases to keep me thinking? A great author, genuine thought leader, and friend!
fooled-by-randomnessFooled by Randomness by Nassim Taleb
amazon.comSimilar to Nate Silver’s book, Nassim discusses how easy it is for people (even professionals) to associate data (i.e. correlate it), when the chances are we are just observing luck. And that correlation does not necessarily mean causality – a mistake so often made in life. It is a good and worthy read, and certainly makes you think. Be aware that this is very much a personal memoir – with the author’s personality writ large throughout.
cartoon-guideA Cartoon Guide to Statistics by Larry Gonick
amazon.comI am a big fan of Larry Gonick’s work – he inspired the illustrations that I use at the start of each chapter for Successful Analytics. Larry has a great pedagogical skill at taking dry subjects (his other books include physics, chemistry, history of the world, etc…) and make them both interesting and directly applicable to the reader in a light hearted way. For example, you can work through the statistics examples yourself and apply them immediately to every day calculations. I studied A-level statistics at school (up to 18 years old) and although not a statistician, I have relied on an appreciation of statistical reasoning to underpin my work ever since. I found this book a great revision on Normal/Gaussian distributions.This is a book to refer to if you occasionally need to dip into standard statistical methods for your day job. Don’t be misled by the cartoon style – his books are aimed at adults looking for a light read/revision of complex subjects, not children. Larry’s background (he studied mathematics at Harvard University) means his books are so, so much better than the “Dummies Guide To…” series!
googledGoogled: The End of the World As We Know It by Ken Auletta
amazon.comThis is quite a different book about Google than you might expect – and refreshing for being so. As an alternative to the plethora of books about the Googleplex, practices, processes, structure, tools and so forth, this book is a history lesson about the men at the top, the ground breaking decisions they were making and their impact (disruption) on so many industries. I am of course bias from working at Google around this period (2005-8), but I found the story fascinating. The emphasis being on the “story” Auletta tells. This is not an encyclopaedia of facts (though there are many), rather a good read about the people being fortunate enough to be part of such a phenomenon.
krug.jpgDon’t Make Me Think! by Steve Krug
amazon.comThis was a seminal book for me when I read the first edition circa 2003. Apart form hitting a chord of mine in terms of thinking about what makes a website successful (or not), Steve writes eloquently about complex web design issues in such clear and simple terms. The book is humorous and very easy to read (less than 200 pages). Unknowingly to him, Steve Krug (pronounced Kroog) inspired me to start writing. The principals described in “Don’t make me think” still influence my thinking to this day.
information-is-beautifulInformation is Beautiful by David McCandless
amazon.comA beautiful coffee table book for the data enthusiast, this is an excellent book for inspiration. Its full of example images and charts though little text in terms of background or methodology. That is both a plus and a minus, but for this type of book a definite plus. The topics are varied with some very relevant and thought provoking charts (cancer survival rates, Iraqi war budget) and some trivial (six degrees of Kevin Bacon, the evolution of pop). However all are imaginative and stimulate the data visualisation debate – that is, how best to show complex data in an at-a-glance way so the important points are remembered.
facts-are-sacredFacts are Sacred by Simon Rogers
amazon.comThis is another excellent coffee table book (beware if you come to my house!) that I bought to compliment Information is Beautiful. However, this one is different – more of a study of the modern day use of data in journalism, rather than how data itself is used. And there lies a slight problem. Simon Roger’s views and his natural bias (political viewpoint) does show through. I felt I was reading the Guardian newspaper (he launched and edited guardian.co.uk/data), when my expectation was a treatise. That said, this book does make you stop and think about how to convey complex data to an audience – and how the same data can be told in difference ways to show a positive or negative perspective. That makes it a welcome addition to my bookshelf!
Purple Cow by Seth Godin
amazon.comA very easy to read and engaging book (you can literally read it on a plane trip!). However, that doesn’t detract from the skill that Seth has at illustrating, with great case studies and in very simple terms, how the world has changed since the web came along. An inspiring read even for people who have seen it all…
Econsultancy.comNot a book, but I am a big fan of Econsultancy. The quality of their industry research reports, and the people that produce them, is excellent. A great resource for anyone working in digital. I regularly refer to their reports and surveys in my books and presentations.
Copyright Brian Clifton