The Canadian branch of the non-profit Search Engine Marketing Professionals Organization (SEMPO) held a seminar in both Toronto and Montreal recently, as a way of bringing digital marketing professionals together to discuss the finer points of their trade. With presentations by Andrew Yang of Bing, and Elena Klingbo of Google, the big search engines were well-represented. But it was Acquisio's Brian Minor who managed to break this otherwise dark science down into comprehensible bite-sized pieces. The man, it turns out, is a rocket scientist. For the layperson, attending a meeting of search engine marketers is a bit like talking to a bunch of people who can see the Matrix all day long, while you and everyone else get by basically appreciating the surface appearance of things, never really understanding the zillions of tiny data points and algorithms that underpin digital behaviour. So getting a glimpse into this world is a valuable opportunity to listen to these people explaining their dark arts, even if the learning curve for an outsider can be a little steep. To begin proceedings Andrew Yang, representing Microsoft Canada's Bing search engine, led with a demonstration of Bing's effectiveness by Googling his own name. It was an effective and witty way to kick things off. One surprising fact learned during Yang's talk is that when an ad is bought on the Bing network, it shows up on both the Yahoo and Bing search networks, which rather surprisingly accounts for 10% of all search queries in Canada. So marketers ignore Bing at their peril, unless they believe that this tenth of the audience counts for nothing. Meanwhile, Google's talk focused on the parallels with and differences between the online and off-line identities of a theoretical businesses. While both presentations by the megacorps were enlightening, they were also perfunctory, amounting to a kind of "This is our solution. Check out the dashboard and what it can do. Thank you for listening," type of run-through. That is why Brian Minor's presentation stood out so drastically. It's an almost meaningless catchphrase to say that something "isn't rocket science" when implying that any task or job is basically not all that difficult when compared to the most difficult job of all: rocket science. So it draws you up short when you discover that the person speaking in front of you is actually a qualified rocket scientist. The only puzzler remaining is why a rocket scientist would bring their skills to bear on the comparatively lower-stakes world of internet marketing. Hearing it laid out in such stark, granular fashion by a man who clearly knows the nitty algorithmic gritty of what he is actually talking about was enlightening, to say the least. The substance of Minor's talk focused on the no-nonsense task of extracting the most value from your advertising campaign for the least amount of money. This is surprisingly easy to screw up, not to mention that for most of us, the means of running an ad spend online is largely an automated process that we barely understand and blindly entrust to the established players in the field. We're making a cruise missile, not a cannonball. Why do cruise missiles hit their targets from 2,000 miles away? Because they steer constantly. It's the same thing we're dealing with here. We're dealing with highly non-linear problems, sparse data sets, and we're trying to make good decisions. - Acquisio Chief Scientist, Brian Minor While his talk started out quite ordinarily, the presentation became interesting when Minor compared most people's bid process (the act of determining how much you're willing to spend to serve ads on a specific platform and how frequently the bid gets updated during a given time period) with how that process works on the Acquisio platform. "You're in the auction room all day," he said. "You're never not in the room. You're there, competing for every single bid. If you use standard delivery, it means you're not there very much. You're only there when Google chooses to put you there." Because search engine marketing (and the general importance of Big Data, A\/B testing, the challenges around placing ads in search, etc.) is still underappreciated by most marketers and business people, they tend to like a "set it and forget it" approach. Or at best, they decide that once a day is an appropriate frequency for bid updating. That approach clearly riles Minor. He compares the once-a-day method to trying to hit a target with a cannonball. To begin with, you're unlikely to hit the target. Even more frustratingly, there's no way to guide the cannonball mid-flight. "There's no steering between when it goes bang and hits the target," he says. "We use something different. What we're doing is what we call 'continuous SEM optimization'. That is, we're constantly deciding, with the best information we have available at this moment, how to steer it towards a solution. So, many times a day, we're going to change your bid. Why does this matter? Because we're making a cruise missile, not a cannonball. Why do cruise missiles hit their targets from 2,000 miles away? Because they steer constantly. They can go right through the window of a building from literally 2,000 miles away. They can put two of them in directly, one after another. It's the same thing we're dealing with here. We're dealing with highly non-linear problems, sparse data sets, and we're trying to make good decisions." Personally, when I hear someone start talking about cruise missiles during a presentation about search engine marketing, I tend to think that this is very likely someone who has watched Top Gun a couple too many times. You tend to keep a closer eye on these people than you do on everyone else. And maybe figure out where the room's exits are, just in case. As I was to learn later, however, Brian Minor has a PhD in Physics from the Air Force Institute of Technology. He worked with DARPA (the Defense Advanced Research Project Agency) on something called "space physics nuclear threat mitigation". He is, in fact, a rocket scientist. Suddenly, I'm listening very much more closely than I tend to do at these industry slideshows. It feels like serious stuff. Taking a poll of agency people in the room, Minor asked, "Raise your hand if you have clients who don't care what you spend per month." One person raised his hand to general laughter in the room. "We've got one. They're like unicorns. They do exist, but they're rare." For the absolute majority of digital marketing people, who are not unicorns, their clients tend to care very much about return-on-investment and getting the absolute most of their ad spend. If you use standard delivery, it means you're not there very much. You're only there when Google chooses to put you there. Advertising has not evolved very much, until very recently, from the days of early-20th century ad man John Wanamaker, who famously said, "Half the money I spend on advertising is being wasted; the trouble is I don't know which half." In addition to focusing on conversions (the moment a person clicks "buy" when presented with the opportunity), Minor's formula involves determining how much each conversion is "worth" (aside from its immediate "buy" value). Not all conversions are created equal. He also talked about differentiating mobile results from home computers. "We're seeing clients now where 90% of their volume is coming through a mobile device." The days of Wanamaker's "I don't know which half" quip are over. Digital marketing is trying to close the gap on making that ad spend as close to 100% effective as possible. The future of advertising is looking a whole lot like rocket science.