When Big Data Makes Life More Convenient

We’re all wary of big data (for good reason) but we have to admit: it’s ushered in technological advances that make our lives easier. I have two clear examples, both Google products, that show what we’re getting in exchange for handing over our data.

The first is the rise of suggested auto-replies in various messaging platforms. I’ve seen these on the native text messaging app on my Google Pixel smartphone, Smart Reply in my Google Business Suite Gmail email draft windows, and even when replying to messages on my Apple Watch. On all three of these platforms, when I get a message, the service will propose three quick possible replies it thinks I could use based on the content of the incoming message. All of these are useful when I am on the go and need to acknowledge a message and don’t have the time to compose something original.

Google reports that Smart Reply has leveraged big data to make this possible with deep neural networks. It not only gathered nuanced responses based on the huge numbers of emails that went into formulating the proposed responses, but also collects data on how each user communicates to refine the suggestions that it proposes to be more aligned with how they write. Google used machine learning to better understand the hierarchy in language so it would be able to make the “subtle distinctions” necessary to propose responses that fit the situation and that users would actually choose to send (Strope, Kurzweil, 2017).

The second technological advancement thanks to big data that I really appreciate is the YouTube algorithm. YouTube has figured out the type of content that I like to watch and curated a feed very effectively, even though I don’t subscribe to many YouTube channels.

When I access YouTube on a computer where I’m not logged in, the recommendations are very general and not at all what I enjoy to watch, so I can really see how much it really has narrowed down.

On the YouTube app, they also have created categories at the top where I can touch to narrow down into my interests. It has provided categories for cars, makeup, culinary arts, a specific video game I enjoy, and a category of video games I enjoy. So when I have sort of an idea about the type of content I want to watch, or need to be prompted, I can narrow it down quickly.

In 2018, YouTube’s Chief Product Officer reported that the 70% of the time spent watching YouTube is driven by their algorithm’s recommendations. YouTube uses big data to tailor their algorithm based on the percentage of people that watch a video or ignore it when it’s recommended to them, the view duration and average percentage viewed from all views of a video, the number of likes and dislikes on a video, and regional context, including time of day and language spoken (McLachlan, Cooper, 2023). In short, YouTube collects data on all videos being viewed to determine if it should recommend the video to people in general, then uses the data about the type of content a user engages in to refine their recommendations further.

I remember years ago when I first had access to streaming services and they did not yet have the huge trove of viewer preference data on me and other users. At the time, the recommendations for what to watch would be fairly generic, and I would spend a long time scrolling to find something of interest. Now I spend more time enjoying the content and less time searching and wondering if I will enjoy the content.

And, content creators know the people watching it have a higher chance of connecting with the message. I appreciated the flip side of that when I ran the YouTube channel at Kajabi because I had the analytics that helped me create more engaging content for our leads and customers.

References

Fernández de Lara, C. (2023, June 14). 6 Gmail AI features to help save you time. The Keyword. Retrieved August 23, 2023, from https://blog.google/products/gmail/gmail-ai-features/

McLachlan, S., & Cooper, P. (2023, April 18). How the YouTube Algorithm Works in 2023: The Complete Guide. Hootsuite. Retrieved August 23, 2023, from https://blog.hootsuite.com/how-the-youtube-algorithm-works/

Solsman, J. E. (2018, January 10). YouTube's AI is the puppet master over most of what you watch. CNET. Retrieved August 23, 2023, from https://www.cnet.com/tech/services-and-software/youtube-ces-2018-neal-mohan/

Strope, B., & Kurzweil, R. (2017, May 17). Efficient Smart Reply, now for Gmail. Google Research Blog. Retrieved August 23, 2023, from https://ai.googleblog.com/2017/05/efficient-smart-reply-now-for-gmail.html