Data Mining and Advertising: What’s Next?

Advertising has long relied on data to make its pitches, refine its pitches, and analyze the effect on sales. In that sense, data mining — “the analysis of raw data to try to find useful patterns and trends,” in the words of Business News Daily — isn’t anything new to the advertising business.

But the vast quantities of granular data now available do make technology news. They can move advertising into new directions than can affect the business strategy of both advertisers and the organizations they serve.

The Second Screen Effect

A recent study performed by a researcher at the University of Pennsylvania’s Wharton School of Business in conjunction with other media analysts, for example, explored the phenomenon of digital searching on large search engines to see whether television ads had an influence on this type of searching.

It found that an influence did exist, particularly with viewers who were using “second screens.” “Second screening” is the term for people watching both television and another screen — mobile phones, pads, or even personal computers — as they watch television.

The study looked particularly at whether the phenomenon of second screens affected whether the television viewer clicked on sponsored ads that came up in their television advertising related search.

The results? There was more of a tendency to click on sponsored ads depending on the size of the screen. Television viewers with smartphones did tend to click on the sponsored ads they encountered in a search relating to the advertising they were seeing on the television screen. But, if the device was larger — a pad or a laptop, say — there was no effect.

The study found that the first sponsored ad in a search was the one clicked, which they attributed to the design of mobile phones, where only the first ad may be seen on a given screen.

Data mining can help advertising refine its message and its audiences.

Future Paths

The ability to find this out opens several potential paths.

First, the ability to follow the linkage with sponsored ad clicking may open an opportunity to analyze the data further to find out where viewers are in the sales funnel. Were they just exploring the product on the second screen? Were they prepared to buy? Were they comparing the advertised product to another product? Once these questions are refined, the advertising can be refined to appeal to different sales funnel positions.

Intriguingly, it will also be possible to add “sales funnel position” to the host of demographic variables known about a viewer, like age and gender.

Second, advertising companies can run ads and then refine them based on the number of clicks on sponsored sites.

Finally, the ability to mine aggregate data could also lead to “addressable television.” Addressable television is advertising or other messages not created for a demographic group, but for the person being streamed a particular show. This is possible currently on streaming subscription services such as Netflix and Hulu because the companies know specifically who their subscribers are and have data on them. Addressable television opens the way to one-on-one advertising.

While advertising companies have always mined data, the vast amounts now available will allow more finely honed messages.