Analytics is helping at least one newspaper publisher to sell advertising on its Web pages.
You want big-data? How about the 6 billion ad impressions that The Financial Times's FT.com Website generates each year, or 5 million a day. Factor in the growing use of mobile devices to view content on FT.com -- now around 20 percent of total viewing, up from maybe 2 percent 3 years ago -- along with a truly global audience hitting the site 24 hours a day -- and the job of figuring out which impressions to sell to which advertisers becomes a serious challenge.
Not big enough for you? OK, add in the fact that these ads are auctioned off in real-time, with three or four advertisers bidding for each one.
The job of overseeing the tricky task of matching the supply of ads to demand from paying advertisers as profitably as possible falls to Jon Slade, global online and strategic advertising sales director at FT.com. And naturally enough, he has been turning to big-data analytics techniques with, he reports, no small amount of success.
In the past, it would take anywhere from eight working hours to two full days for his operations people to obtain an ad inventory forecast. Now that job gets done in eight seconds. By freeing people to focus on more strategic issues, analytics has enabled FT.com to achieve double-digit growth in ad yields and a 15 to 20 percent improvement in the accuracy of metrics for its ad impression supply.
Such are just some of the intriguing findings available in the latest issue, just out, of Technology Forecast, a quarterly publication prepared by PriceWaterhouse Coopers. It's available for downloading at no charge from the PwC Website, and I can highly recommend it for anyone who's looking for a lucid, in-depth discussion of the current wave of analytics technology and what it really means to business.
The FT's Slade happens to be one of several executives at user companies and big-data analytics providers whom PwC's Center for Innovation and Technology has interviewed for its well written, graphically enticing journal. Slade tells the PwC team how The FT has been using analytics to understand its readers better and thereby to give advertisers a better return on their investment:
We examine user actions, and we note that people who demonstrate this type of behavior tend to go on and do this type of thing later in the month or the week or the session or whatever it might be. If we know, for example, that people type A-1-7 tend to read companies' pages between 8 a.m. and 10 a.m. and they go on to personal finance at lunchtime, then we can start to examine those groups and drive the right type of content toward them more specifically.
Helping with all this, Slade points out, has been a company called Metamarkets, a startup that describes itself as providing data-science-as-a-service (DSaaS?). It has developed its own cloud-based, in-memory database engine, called Druid, for high-speed processing of massive datasets. An interview with Mike Driscoll, CEO of Metamarkets, also appears in the Technology Forecast issue. Among other insights, Driscoll offers this:
Many businesses keep only 60 days' worth of data. The storage cost is so minimal! Why would you throw it away? This is the shift at the big-data layer; when these companies store data, they store it in a very expensive relational database. There needs to be different temperatures of data, and companies need to put different values on the data -- whether it's hot or cold, whether it's active. Most companies have only one temperature: they either keep it hot in a database, or they don't keep it at all.
There is much more of this kind of thing in the PwC journal, including a run-down of four critical big-data technologies: in-memory technology, interactive visualization, statistical rigor, and associative search. And one article is devoted to natural-language processing, which is proving useful in mining social media data.
My advice: Get this new issue of Technology Forecast and read it. You're bound to learn something.