Dr. Usama Fayyad the CEO of BI consultancy Open Insights has an impressive resume. Until recently Fayyad served as chief data officer for Yahoo! Prior to that, Fayyad co-founded and led the DMX Group, a data mining and data strategy consulting and technology company. He also lead the data mining research group at Microsoft. At today’s Predictive Analytics World conference, Fayyad took the stage and shared details about his experience with data mining and marketing, specifically at Yahoo!.
“I thought being someone who worked at Microsoft and the Internet in the early days and started an analytics firm and ran a data mining company, I should understand the Internet,” Fayyad told the crowd. “I realized [at Yahoo!] there was so much I didn’t know” What we needed at Yahoo! was data strategy, said Fayyad — strategy about how to make data a driving asset and how you can dictate new business from it.
“When I joined Yahoo!, all I understood was marketing from the top to make people aware that you have a product and trying to reinforce your brand. Then, at the bottom, is marketing before a purchase.” He relayed that essentially Yahoo!’s business comes down to targeted advertising for brand awareness and direct marketing to capture customers at the cusp of buying.
Knowing this, Fayyad began to ask, what’s in between the two? Where are the marketers when customers are in the “consideration” phase? “The zone in between is more interesting and more profitable,” he said and went on to talk about search engine marketing and the art behind companies “bidding” on search results.
As search engine optimization evolved, the concept of behavioral targeting was brought to light, requiring more advanced analytics than ever before. The challenge companies were facing was: I have customers I like or who are interested, but I want to find similar customers who haven’t shown interest yet. Enter Yahoo! and its database matching capabilities.
“This stuff really works,” said Fayyad about behavioral targeting. The science behind it is matching a Yahoo! users’ profile with client’s customer database. Through profiling, data reveals unique behavior that is easier to target. Fayyad reveals stats attributed to behavioral targeting:
*66 percent higher interest of consumers
*Behavioral targeted users were 29 percent more likely to click on an advertisement.
*Organizations, on average, when behavioral targeting customers increased revenue by a 200 percent lift.
Now behavioral targeting is evolving to involve social media and a user’s social activity. Fayyad talked about his fascination with Flickr (a site that Yahoo! acquired in 2005). “I was very wrong about this at Yahoo!,” Fayyad admitted, saying his reaction to the site that allows you to put up pictures, share them and let anyone comment was, “‘Really? Who would want to do that?’” He went on, “Jerry Yang said, ‘This works.’ — He was right, I was wrong.”
The basic idea, Fayyad told attendees, is we can gather information about social relatedness. Marketers can find ads that users tend to click on then find the users in that person’s social network to target ads to them, as well. Fayyad said Yahoo! experimented with this in the categories of weight loss, food and nutrition, and fantasy football– and the results of marketing to a user’s social circle, were significant. “You get this amazing lift by going off the extended population,” he said. Friends tend to think — and buy — alike.
Fayyad, although now departed from Yahoo!, said that the doors are open for the next generation of online marketing. Targeting social media is just beginning. “Online marketing is ripe with opportunities for predictive analytics applications,” he said. “And the results are very powerful.”
Yet, Fayyad admitted to not knowing the answers of what the future will entail. “As big as the [search engine] industry is – it’s a $25 billion industry – we still haven’t figured out a lot of the basics.”
By: Lauren McKay