Published by Crown on September 9th 2014
Buy on Amazon
I received this book for free from NetGalley in exchange for an honest review. This does not affect my opinion of the book or the content of my review.
This post contains affiliate links you can use to purchase products. If you buy using that link, I will receive a small commission from the sale. It will never cost you any more.
Did you hear? Dating site, OKCupid has lied to you; just to see what happens. This headline hit just days after Facebook tried publishing user behavior research in an academic journal. What the journalists seem to have missed was that OKCupid’s co-founder and President Christian Rudder wrote the blog post about some of their findings just before his book, Dataclysm was released. I actually suspect Christian wrote the blog post because of huge backlash Facebook received. After all, a blog post worth of user behavior data is easier to stomach than an entire book.
So what’s Dataclysm like?
It’s an informative, educational look at people and what they do. Is it a scandalous expose? Not really. Will you be surprised by the results? Probably.
Christian takes what comes across as a math nerd’s hobby and turns it into an insightful profile. He has access to gigabytes of offered and acquired data. I know I wouldn’t be able to resist.
I suspect this book has two aims. One, to show what data is available for analysis, and two, to research some behaviors that are difficult to accurately measure. For instance, do men search for gay porn more in liberal states? By the way, no they don’t. Search rates are equal across the country.
Other little snippets are reported from data that extends to Google, Twitter, a job site and more. Academic research also supplements the OKCupid sample, giving a more robust story than just that from a dating site. Some snippets are useful for marketers, such as the fact that people are more likely to reword a Tweet than use abbreviations. However, most of the data is general and an interesting anthropological view.
Christian’s story telling tends to be more pop sociology with simplified English. He does drop just enough research terminology to keep the data nerds happy, but always with translations. Chapter titles like, “Death by a Thousand Mehs” helps grab those who detest math.
The book could be tightened a little with some setting the scene paragraphs being dropped. I do especially like the “end of book philosophical chapter”* that explains how web data analysis is here and should be useful for consumers, but of course needs to be treated cautiously. He quotes the Target case where their data modeling was so accurate they predicted a pregnancy before the woman told her family. Unfortunately the woman was a teen. He’s right though, data analysis is here and really we should embrace it.
Who Is Dataclysm For?
Dataclysm is more of a sociology book than a marketing book. If you’re a marketer wanting to understand the applications of big data, then definitely read this. It won’t help a marketer do their job better. If you’re worried about online privacy and want to understand what is recorded, then definitely read this book. Finally, if you’re just a curious nerd, buy it. My copy was an unedited proof courtesy of NetGalley, without the graph formatting. I now have to wait until it’s released next week to buy a full copy.