Is Big Data Enough? A Summary of a Ted Talk by Tricia Wang and my Thoughts

    

Hello Everyone!

    I recently watched another Ted Talk titled “The human insights missing from big data,” by Tricia Wang, and wanted to write about what I thought about it. In her talk, as the title suggests, she discusses how big data cannot do justice to the task of predicting patterns, and that human insight (or what she calls "thick data") combined with quantifiable data will elevate the performance of companies and nonprofits, making them more profitable and more robust in the face of the future. 

What Wang Discusses

In 2009, Wang took up a research role at Nokia, when the company was dominating the emerging markets of the cellphone industry. Wang talks about her experience in an urban slum in China, where she spent years selling dumplings to construction workers, learning about the kids' habits, and gathering quantifiable data on the population. Piecing together all she had learned, she discovered something the wide array of Nokia's big data couldn't predict: even in poverty, these people wanted iPhones. Wang describes trying to convey her first-hand "thick data" to Nokia, which discarded her insights as unsupported by their big data algorithms. Nowhere in their models was there a strong positive sentiment or desire for iPhones. 

    And so what happened? Nokia went from dominating the cellphone industry to becoming obsolete. Wang indirectly suggests that their downfall stemmed from their inability to incorporate the small but powerful human stories of technology ethnographers like herself. She points out that over 73% of big data projects fail to deliver profits or breakthroughs, largely due to internal quantification bias that can blind companies to unmeasurable insights. She highlights that this doesn't happen only at Nokia but also in policing, where data is used to predict crime, reinforcing past biases.

    Although it may seem she is criticizing big data, she instead argues that big data isn’t wrong; it just isn’t enough by itself.

Wang's Philosophy

Big Data Algorithms + "Thick Data" = ↑ understanding of customers and trends

What I Think

I agree with everything Wang says in her TED Talk. Its definetly vital to implement the insights of those who have experienced the unquantifiable. However, I wonder how exactly companies do this: how do Wang and other ethnographers turn their insights into data that can be fed into models similar to those used for big data? Wang describes how Netflix incorporated the "thick data" of ethnographers into its big data after identifying that consumers loved binge-watching. However, how did they turn that lived experience into quantifiable data to use alongside their big data models? Seems like a very intricate process that I assume both ethnographers and data analysts have to handle. Additionally, this TED Talk is from 2016, so I wonder how Wang's perspective has changed since then. I wonder if big data models have gotten so good at analyzing past patterns to predict the future that "thick data" is less relevant. Or maybe industries have identified the importance of "thick data" and now always include it in their big data predictions.

If you want to check out the TED Talk where all of this came from, here it is: https://www.ted.com/talks/tricia_wang_the_human_insights_missing_from_big_data?utm_campaign=tedspread&utm_medium=referral&utm_source=tedcomshare