As we watched the Iowa caucus coverage this week, our digital marketing brains started thinking how difficult it must be to manage digital marketing for the Trump brand. All those properties, buildings, golf courses, resorts, and products out on the market, and then you throw politics into the mix. The team must have a lot of noise to sort through every day in order to connect the data, and it can’t be easy to stay on top of what people are saying about the Trump brand, and candidate Trump.
Parsing out data is major challenge for businesses today as everyone attempts to become data driven in developing their strategy forward. Our high velocity, high volume, online, social, enterprise, and device data-driven world requires a foundational infrastructure to separate the signal from the noise in order to discover valuable insights and make things click. And, although there are many approaches to gleaning insight from high velocity data, many overlook a simple and proven solution — taxonomy.
Arriving at inspired moments where you achieve brand engagement takes a great deal of effort, and brands like Trump’s are constantly developing new approaches to working with data to create hundreds and thousands of inspired connections each day.
Where do we fit in? We support the infrastructure for data-driven efforts as the primary filter for high velocity social data streams.
The social media challenge
Social media is the great connector these days, and the data produced by social media is highly valuable. Politicians like Trump will rely heavily on social media channels during this election cycle. But social data is also difficult to harness because it is short form in nature. Traditional means of text analytics can’t interpret short snippets without a large corpus of data on which to train.
Social data also comes at such a high velocity and in such great variety — more than 500 million tweets per day — that it’s difficult to categorize all the data, particularly in real time. The nature of social media makes more traditional natural language processing (NLP) tools stutter and stop when confronted with slang and abbreviations. Deciphering nuance is overwhelming for machines.
This leaves digital marketing teams with the problem of trying to connect moment-to-moment individual behavior with the larger themes in order to establish better connections.
Donald Trump’s team, for example, has a few problems to parse through every day. The Trump brand has so many inferences — hotels, casinos, golf courses, products, and now, politics — so the digital teams on the brand side and on the candidate side need powerful tools to sort out all the noise on a daily, if not hourly, basis as Trump the candidate continues on his journey through the primaries. Once they’ve separated the politics from the rest of the Trump empire, they still have a lot to do to drill down through news, facts, opinions, and chatter that they want to study in order to make better digital and social decisions on behalf of the brand.
Connecting the data: a problem taxonomy can solve
Brands that use eContext’s deep taxonomy to drill down into the data can pull out the connections to understand what their fans are thinking and doing.
Our taxonomy extends 21 tiers deep and features 450,000 total categories, along with 55,000 language rules. It’s a complex taxonomy that helps brands more accurately measure the distance between concepts. Using eContext, Trump’s various digital teams can cut through the noise to find out that “Trump views of Chicago” are about his property in downtown Chicago, rather than his thoughts on Windy City politics.
eContext is the largest text classification engine available. We help brands structure data from any source, running it through our universal topic hierarchy in real time.
Our taxonomy is curated by people. We haven’t put our faith solely in machine-learned associations or the edges and weights of a graph. Instead, we know what a thing is because of where we place it in our hierarchy, and we make sure that we know for certain what it is rather than relying on other data and extrapolating.
Even the most elusive or rare data has a defined space in our structure, and there’s never a case where there is insufficient data to describe what something is.
Brands put our technology to the test constantly — using data from a wide range of sources, and applying it in creative ways to help their roster of clients arrive at better decisions and outcomes.
The end result of all of this high-tech data crunching is focused on the very simple idea of helping all of us connect with the content that we want to see, hear, or watch. This is going to become more important as brands seek smarter ways to connect with people. And let’s face it, a brand like Trump doesn’t have 30 days to wait for data to be refreshed – the team needs answers in real-time to help them make fast decisions on the campaign trail.
Getting things to “click” is a major challenge for businesses today as everyone attempts to become data driven in developing and driving their strategy forward. Our high velocity, high volume, online, social, enterprise, and device data-driven world requires a foundational infrastructure to separate the signal from the noise in order to discover valuable insights and make things click. And, although there are many approaches to gleaning insight from high velocity data, many overlook a simple and proven solution – taxonomy.
We, at eContext, specialize in taxonomy, and the applications that add value to a myriad of use cases, including data storage and findability/discovery, digital advertising, recommendation engines, consumer insights, and competitive intelligence. And today, we are happy to be writing a guest blog spot with Black Swan on our expertise of taxonomy as it applies to a channel growing incessantly more complex every day – Social Media.