Stop Hiding Behind Your Algorithm

Alan Pearlstein In the past couple of years, a new display ecosystem sprung up rapidly, this time built around the separation of media and audience. Every week it seems as if more data is available, a new tech provider enters the market, and another snappy startup claims it has some kick-ass algorithm that will change the world for marketers. The business is exploding so rapidly that even those on the inside need slides and charts to figure where every company fits in the landscape. Every new company promises that their algorithm delivers better results, better targeting, and utilizes better audience data. These new enterprises spring up to fulfill a market need and claim they can solve a very complex industry problem. Trouble is, most of them are full of bull. In the end, these companies never come through on the promises they make, and the market is left looking for a real solution. We hear from clients all the time that the new businesses soliciting them are confusing, and no one understands what they claim they're doing. On top of that, most buyers and brand advertisers are a lot savvier than we in digital give them credit for. Young companies, take note: media buyers see the word "algorithm" and they have questions. At this stage of the game, a good media buyer knows that some companies hide behind the word when they can't show an actionable model. It isn't enough to just say you have a fancy algorithm – you have to be able to explain how and why it works. Many algorithms do work well and can actually match the right advertisements with the right consumers, aggregate new types of audiences and more. But sometimes the "proprietary algorithm" doesn't really exist. It's just tech slang that hides the fact that the company can't provide any results or a differentiated product. Some of these companies actually resort to shady tactics to mask the fact that their algorithm does nothing. For example, there are a number of companies in our space that claim to have created algorithms that can auto optimize an ad campaign to drive efficient customer acquisition. They convince the marketer to give them too much credit for view conversions, and then the real scam begins. The companies buy low price, below the fold, garbage inventory so that as many browsers as possible have their cookies (it's not hard to reach the majority of the online audience when you are buying $.10 CPM inventory). Then they take credit for every sale that occurs from the audiences they have sprayed with ads – even though the vast majority of the users never even saw the ad. The advertiser, meanwhile, has no idea that the inventory is well below the fold where no consumer will ever click or engage. There's no magical algorithm or data-driven decision-making involved here – just a company pulling the wool over the eyes of their clients. Buyers are learning to see through this. They're no longer buying the overcomplicated tech speak and lingo, and they're demanding transparency. If your pitch involves the phrase "just trust me," good luck winning any deals. Nobody in this business wants to give away their secret sauce and show the minute details of their algorithm, and that's fair. But it's imperative that you can walk into a room with a media planner and actually explain what your system does, how it works and where the data comes from. As my friend and colleague Darren Herman tweeted recently, "Algos are only as good as the data they sit on top of. My fantasy football teams reminded me of that yesterday. #gotcrushed" The reality is that most algorithms are worthless, not because the logic of the algorithm is flawed, but because the data used in the algorithm is inaccurate. Garbage in, garbage out; it doesn't matter how smart the algorithm is. Companies that are driven by algorithms need to put more focus on the quality of the information and data driving their product. One red flag is when the company won't tell you where the data comes from. There is usually a good reason why they don't tell. Media buyers want to use the newest technology to reach the best consumers with innovative campaigns, but some concerns outweigh the desire to be an early adopter. Media buyers are looking for real products and insights, not magic. The winners in the new data/media space will be the companies that can answer questions from media buyers and develop the managed services and products that are understandable to buyers – not cloaked in fancy lingo. ABOUT THE AUTHOR Alan Pearlstein is president-CEO of Cross Pixel Media.

Stop Hiding Behind Your Algorithm

Featured Stories