Today’s blog comes to us from Andrew Panos. Andrew has been in social analytics for nine years, and currently works as Essence Digital’s social media researcher. He also thinks Mad Max was the best movie of 2015, and is prepared to argue about it in the comments section.
The anticipation is coming to fruition. All the hard work has come down to this. Glory is gained. No, it’s earned here.
As you know, this year at the Academy Awards, the category of “Best Sound Mixing” will be a fist fight. First-time nominees Scott Gordon (Mad Max), Wayne Elliot Knight (Chunnel) and Geoffrey Trent (The Revenant) will face off against past nominees Harry Lockhart (Ex Machina) and Stephanie McGillis (Bridge of Spies). Fasten your seatbelts!
Now, you might be saying to yourself, “What the hell is sound mixing? I don’t care about sound mixing.” Two things. First, sound mixing is essentially the act of composing sounds in a movie, whereas sound editing is the process of making those sounds. Second, of course you don’t care about sound mixing. Heck, the names above were all made up, and Chunnel is a fake movie referenced in Seinfeld.
But, if you watch the Academy Awards, maybe you do kind of care about sound mixing. Or, the Oscar for “Best Animated Short Film.” Or, “Best Makeup & Hairstyling.” Whatever the category may be, whether you’re watching alone or with a group of people, when you hear, “…and the nominees are…” you try and guess who will win. We all do it; we want the sense of being right for a second, or having that “I knew it moment.” More so, you may be in a prediction pool where there’s a prize on the line.
Predictions are enjoyable for things like the Academy Awards, the Grammys, and the Super Bowl. Observing whether winners correlate to social media mentions, or “likes,” is a fun thing to look at, too. For some of us (hi there!) it’s our job to make predictions based on social media.
Using social media analytics tools, like Sysomos in this case, we researchers go through musings and ramblings, compare past trends, and make predictions on how/when/why people will talk online. Results then inform our managers and clients; it’s really forecasting as much as it is predicting.
For instance, mentions of “food” drop significantly on Halloween. People ask for happy hour recommendations most often on Fridays at 4PM. “Today” mentions jump 50% on Valentine’s Day because, fellas, it’s a very important day. We make our trend predictions based on moments like these and tailor work around them.
Showing when a brand or campaign is talked about and basing forecasting around these mentions is smart, but sometimes not super helpful. A good social media forecast gets curious; it dives into random thoughts that might pop into the researcher’s head. If you’re looking at Walking Dead mentions, you can see when people talk about the show on Sundays, but curiosity drives better reporting. Perhaps it would be more interesting to pinpoint the moments Daryl is mentioned by people with over 10,000 followers during the show .This can drive content on social platforms during the next week’s airing, and it even comes with a glance at influencers.
Back to The Awards…
All that being said about curiosity, we’ll get back to using social media to predict award show wins. Articles on this topic, like this one you’re reading now, usually run before airings. The common theme is: This person had more mentions, therefore social media says they will win an Oscar. These types of “predictions” are fine, but researchers can do better; they can get more curious. Below, I’m going to make predictions on who will win at the Oscars and compare those picks with who oddsmakers from Easy Odds, and experts from Variety and Vanity Fair believe will win in all of the categories. How will this be done? By sound mixing.
Given that singular mentions can be boring, and sometimes inaccurate, how can we make the forecasting model better? We have to mix it. First, let’s remove all retweets and links. Now, articles and viral posts won’t affect our numbers; we’ll get pure conversations. Then, we’ll make every nominee category specific. A Mad Max mention will only appear in our “Best Picture” forecast if it mentions “Best Picture.” Same rule goes for its “Best Makeup & Hairstyling” nomination. Lastly, we’re not going to look at overall mentions. The Academy Awards are a slog during the weeks leading up to it, with other award shows impacting predicted nominee success and greatly altering total numbers. So rather than total mentions, we’re going to find which nominees’ polynomial trendline is highest at the end of our timeframe.
Charting trends is not as flashy as gambling sites and their odds, or as widely shared as movie experts on large-scale websites. It’s no “Best Picture.” Social media analysis is “Best Sound Mixing.” We take what we have–the sounds of people talking online, the opinions of the many–and we arrange it into a piece of work. This is the kind of work we do for clients, but it makes for a fun Oscar experience.
Best of luck with your guesses. Take a look at predictions below, and enjoy the show!