Spotify thinks my 10 year old daughter is 25. To be fair, Instagram thinks her 12 year old friend is 23. And Apple thinks my 13 year old neighbor is 17. Welcome to the self-reported user data that will wreak havoc with your data strategy. The reality, and we all know this, is that people will say what they need to say in order to achieve the outcome they desire. In this case, every kid I know has lied about their age to some company somewhere in order to gain access to whatever web-based service they want to use.
I know what you should be thinking right now — this is old news. And you’d be right for thinking that. Everyone knows that the demographic data upon which media-buying decisions are made is self-reported. (Everyone knows that right?) And yet… as my daughter informed me of her “age” on Spotify, it struck me that some ad-tech agency or digital marketer somewhere is buying ads targeting “25 year old females living in Los Angeles who love pop music” and they’re getting my 10 year old kid. Which, at least to me, is hysterical… in an ironic way.
Marketing and advertising exist for a reason — generating sales and driving revenue. There’s no question that the ROI for your ad spend, your ad targeting, matters. Whether your goal is keeping existing customers or finding new ones, a data strategy based upon better look-alike modeling for your ad spend is critical to success. And, at least compared to the way we used to buy ads for print or television, we are making great strides in look-alike modeling.
BUT, and this is the big but. The quality of the decisions you make is dependent upon the quality of the data you have. Stated differently, how big is your data? Consider this: only when you can correlate multiple data points can you find the anomalies — like all the kids (and adults) I know of who regularly lie about their age — in order to get more accurate insight into your customers for more effective advertising. By contrast, if you’ve only got limited data available for your modeling and spending decisions, and much (all) of that data is self-reported, is it really of any use at all? Maybe. But then again, it may simply be a case of garbage in, garbage out.
Which brings me back to the risk my “25 year old daughter” poses to your digital marketing strategy. The truth is that someone out there knows that my 10 year old is not 25. Spotify may, in fact, even know. The question is, when you’re making your ad buys, do you know? And if you don’t, how much is your ignorance costing you?