If you’re a media company, or a publisher, and you look at “Big Data”, chances are you’re not talking about something like the 25 Petabytes produced annually by the Large Hadron Collider. The visitors of your sites or the users of your apps may be a large group of unknowns to you; but you’re not trying to find a Higgs in a haystack. You’re likely thinking about a lot of data; but is that really Big Data?
At the same time, though, you’re not running some mom & pop magazine stand where you know all your regular customers by face. If you’re not doing any transactions, you won’t even have customers — you have an audience. Your web analytics are hopefully giving you some idea of the size of this audience. Maybe even some rough demographics. But it’s still more of a large, faceless crowd; nothing like the specific individuals in a CRM system for sales.
For online publishers, the most interesting part is in between the extremes — somewhere between Petabytes of raw data; and “How are you this morning, Mr. Bloem?”. The audience consists of visitors; to know your visitors is to understand your audience. To know your visitors also means you’ll be able to serve them what they want to consume.
Traditional web analytics aren’t great at this — they give you numbers, such as “unique visitors”, and some basic demographics, like where these visits were from. But there’s a lot of guessing involved with anything beyond that (are these people who like to read about cars? are they married? are they about to buy a house? would they be interested in crochet?). Web analytics process a lot of data. But the information they provide beyond the anatomy of visits is not very rich, and it’s definitely not good enough to target content or advertising or to personalise the experience.
If you’re not careful, it may mean your advertisers actually know more about your visitors than you do yourself. They will set third party cookies, and string together web profiles from multiple website visits. (This is somewhat controversial, and Safari blocks the third party cookies.) If your online properties are large enough, you’ll want your own, first party data; and it has to be much richer than what others could skim off the top.
This leaves a bit of a conundrum for publishers. How do you find out who your visitors are? By creating touchpoints. In the absence of actual transactions (such as ecommerce sales), you have to offer value to visitors in exchange for hints about the visitor. Features on the sites and apps (like profiles, playlists, commenting, gamification); competitions and signups (for instance, for newsletters); making this easier on the visitor with social signup (Facebook, Twitter, Google login).
Privacy is an important concern here — Internet users are increasingly wary of what happens with their information. Publishers need to embrace this, rather than ignore or battle it. There’s no need for actual persons to see everything other actual persons are doing online. This is about anonymous profiles and automatically improving user’s experience; or aggregating it into larger segments and relatively generic (yet precise) segments. It shouldn’t be about Big Brother reading your email.
For media and publishers, it’s about research; personalisation; and advertising. You can call it “Big Data” if you want to surf the hype; or call it CRM if you want to maintain the illusion it’s just a large rolodex. The reality is in the vast grey area in between, and there’s a lot to explore there.