Michael Lewis has done it again. His book, Moneyball (2003), was an inspiration to many of us. He and Billy Beane helped us to start thinking differently about how the right analytics model could lead to unique insights. He taught us that we could identify new and often overlooked value if we looked at a wider range of variables. It was a primer for us as we developed algorithms that led to our Meme, Muse, Blueprint and many other important products.
After Moneyball, Lewis taught us about the importance of the left defensive tackle (Blind Side), collateralized debt obligations (he made it sound simple) and several other interesting reads. Now, he is once again changing how we think of the use of data via Flash Boys, which struck me as a sort of Moneyball, Part 2. We just shifted from the baseball field to electronic trading.
Wall Street has a ton of data and they measure trading success in milliseconds. It turns out that by understanding how data is shared, as it relates to trading, it can lead to major upside for those firms who harness technology correctly. And that is what hit home for me. The Street figured out that if they could understand how to create advantage via technology and access data difffently, they could make significantly more money than their peers. The technology advantage was larger than the industry knowledge advantage.
I always used to think that money was made if the stock simply went up or down, based on market information. I never thought about the opportunities inside the tiny windows of trading time within each stock purchase or sale that are largely irrelevant to market information.
And that is what made me think of what we’re now doing and what’s next in analytics.
Our world is increasingly complex due to the crush of available data and the expanding use of technology. We can access data, publicly or privately, from more sources than ever worldwide. Advertising is shifting towards automated buying online. Large social channels give their clients partial, but not full data. Some new technologies, like SnapChat, have billions of interactions, yet don’t even share or accumulate data.
The world of data is and will continue to get larger and more foggy, not more clear, if you are using the basic tools of the past.
The result of this complexity is a new combination of science and art, which is resulting in the emergence of industry and brand specific filters. We used to think of filters as being fairly basic. It’s now time for filters to play a leading role in the world of big data. They are the emerging heroes for brands or will be soon.
Imagine micro-targeting and filters coming together. In the future, a brand will have a suite of filters that we’ll use like the lens on a camera. If we want to see what skateboarders in Chicago think about Nikes, we’ll put on that filter. We may have 50 per brand. We will also improve how we look at causality across these filters. For example, if we run paid media in twitter, where should we measure to see the total impact? We might measure twitter for earned media, Instagram and YouTube, for example. The key is that, based on patterns established over time, we’ll know the right causality metrics to review within the filter that shows the customers we care about. After all, if we get 50,000 views of a video and it doesn’t lead to any meaningful behavior and it is not hitting the majority of our target audience, do we care? It is sort of the modern day equivalent of the tree falling in a forest. In that analogy, we think of silence. In today’s analogy, we often have a lot of noise, but if the noise is irrelevant and no one we care about is listening, does it matter?
Filters and the new metric bundles will become like Lego’s. We’ll be able to add or subtract to them to create the vision required to understand what to do next. Imagine going to a taxonomy for your brand or category, tapping on the items you want to analyze and the filter and metrics are created for you. Over time, you keep refining your suite of filters and the metrics bundles, so that you can scale their use to your teams worldwide. You will have hundreds of analytics options ready to be finely tuned to the needs of your brand.
Meanwhile, the world will continue to swim in too much data, too many APIs and too many technology solutions that we only partially understand.
You will be looking at the market with 20/20 vision and deciding what move you will make next. You’ll be playing chess in a world that often plays checkers.
All the best, Bob