David Schweidel Professor of Marketing

Headshot of David Schweidel 1x1  @dschweidel

Areas of Expertise

  • Social media
  • Political advertising


David Schweidel, PhD, professor of marketing at Goizueta Business School, is an expert in customer relationship management and social media analytics. His research focuses on the development and application of statistical models to understand customer behavior and inform managerial decisions. His work has appeared in leading business journals and garnered numerous awards, and he has been recognized as a leading scholar by the Marketing Science Institute’s Young Scholar and Scholar programs and by Poets and Quant’s “Top 40 Under 40.”

Schweidel is the co-author of Social Media Intelligence (2014), which examines how organizations can leverage social media data to inform their marketing strategies, and author of Profiting from the Data Economy (2015), which looks at the impact of big data on businesses and consumers.


Facebook Live: Georgia - State of Change: Red, Blue, or Purple? (Dec. 2, 2020)

Georgia’s transition from red to blue

I’ve lived in Pennsylvania and Wisconsin, both being states we traditionally think of as battleground states, so I’m used to seeing the high volume of advertising that typically accompanies all the elections. I was a little bit surprised to see it in Georgia. We expected it to be close. I didn’t know how close it was going to be or how much money was going to be pouring into the state, but if you turn on the television today you’re probably going to see wall-to-wall advertisements for the runoffs.

Expectations for the runoffs

Typically, the way we think of political advertising is that it has two purposes. It’s either to persuade people, to get them to see your side your side of things and vote for your candidate, or it’s to get them to show up at the polls. I suspect most people have probably already made up their minds who they’re going to vote for, so I think it’s a question of how effective that advertising is at getting people to show up to vote.

We’re seeing record numbers of absentee ballots being requested for the runoff election, so it’s working as far as maintaining the level of engagement that people have had. I would expect some of the ‘heavy hitters’ from both sides of the aisle coming down to Georgia to support those ‘get out the vote’ efforts, to try and energize those bases to show up and vote.

Looking ahead to 2022 elections

Watch for more about the role social media can play in elections. “There’s a lot of heterogeneity in the thought around should social media companies be regulated and should they be held responsible for the content that’s posted on their platforms. Should they be liable for the content that they are enabling to be disseminated? I think that’s a valid debate that we need to have.

One way of thinking about this is that the social media platforms have become the de facto publishers, they are now the de facto news outlets. I think there’s a debate to be had on whether we need more regulation around social media as far as misinformation and the damage that it can potentially cause. I give the companies credit for their efforts. Combatting misinformation is like a game of whack-a-mole. You spot one thing and are able to tamp that down, but you never know where the next piece of misinformation is going to come from.

View webinar here: https://bit.ly/2LswBjv


This is pretty much it for both campaigns as far as having a captive television audience, so I would look for them both to try to capitalize on it. After Trump’s COVID diagnosis, we have seen him come back to doing rallies and almost downplaying the pandemic based on his rapid recovery. I don’t think that’s going to play well given the spike that we are seeing in numbers at the moment. There may be a concern that Trump is “on tilt”, shooting more from the hip and see if he can land a big hit on Biden based on how close the race is. Lashing out at the moderator before the debate seems to be an attempt to set expectations and have a story/angle for supporters will flock to. For Biden, he seems to have been able to connect with folks. He hasn’t landed a knockout punch, but his style has stood in clear contrast to Trump’s. I would expect him to maintain his calm demeanor, despite the personal and familial attacks that will likely be lobbed during the debate.


We saw clear spikes in conversations surrounding the VP debate and town hall. Biden/Harris generated approximately 750K mentions in its peak hour the night of the VP debate. In contrast, the town hall generated 380K mentions. Trump/Pence generated approximately 400K mentions in its peak hour during the town hall, not much different from Biden/Harris. During the VP debate, the Trump/Pence campaign got 1.5MM mentions. Based on this, there appears to have been more interest in the VP debate than the townhalls. Part of this could be fatigue from the election cycle, part of it may have to do with the format of dueling town halls. These numbers can’t be looked at in isolation, they have to be looked at relative to what’s normal for the campaign to see if they were able to build any momentum. Since the VP debate, mentions of Trump/Pence have come down from over 7MM mentions to ~2.5MM mentions daily, while Biden/Harris from 3.25MM to about 2MM over the same time period. Taking out the spikes associated with the debates, there’s a slight upward trend in mentions of Biden/Harris over the last month. In contrast, mentions of Trump/Pence look to be pretty flat, possibly with a slight increase.

Chart 1 above is the volume of Trump/Pence, with the first debate and VP debate standing out pretty clear.

Chart 2 above shows Biden/Harris. You can see the gradual build that Biden/Harris have had on social media. You’ll also see a spike in Trump/Pence after the COVID diagnosis that doesn’t show up on Biden/Harris. The part that stands out to me from the town hall is that Biden/Harris almost doubled their mentions compared to two days before the town hall. The town hall had virtually no impact on mentions of Trump/Pence. Based on that, one could draw an inference that the town hall had a bigger impact on conversation around Biden.

Trump and Bloomberg Super Bowl Ads - Post Game Analysis (Feb. 3, 2020)

A preliminary look at how Trump vs. Bloomberg faired on Twitter.

  • The volume of tweets mentioning Trump, his twitter ID @rraldonaldtrump or the hashtag #trump today were 3.27 million tweets. This is down from the average daily volume over the previous week of 4.85 million tweets, a decrease of about 1/3.
  • The volume of tweets mentioning Bloomberg, his twitter ID @mikebloomberg or his campaign’s ID @mike2020 were just shy of 400,000 tweets. This is up from the daily average over the previous week of 132,000 tweets, up about 200%.

So, while social media activity mentioning Trump dwarfs that mentioning Bloomberg, Bloomberg did see a bump that we can likely attribute to the Superbowl ad (and the associated press coverage).

 What about the content of those tweets?

  • Bloomberg related tweets are holding steady at approximately 11-12% positive. He did see a jump in negative tweets today compared to the previous week, from 26% to 40%, which may be in part due to the increased visibility from the Super Bowl ad and more people talking about him. Digging deeper into this, there was an increase in prevalence of posts containing sadness, disgust and anger.
  • In comparison, negative sentiment related to Trump has held steady at around 40%, but he did see a jump in positive sentiment compared to the past week’s average, up to 20% from 14%. Underlying this is an increase in the fraction of posts expressing joy.

The data could be spun different ways. Bloomberg’s campaign will obviously be happy that the issue-related ad boosted the volume of relevant twitter activity by as much as it did. As they try to reach as many people as they can, the increased negative sentiment is probably not as much of a concern at the moment. It will be interesting to see if the campaign can take advantage of the ad and sustain the momentum over the next few days. It’s a bit surprising to see that the President didn’t get a bump in Twitter activity following his two ads, but Trump-related twitter activity already is at a high level, so there may be a ceiling to how high it can go.

Of course, we’re only looking at social media, so any data on website traffic and/or donations coming in would give us a better read on how effective the ads were.

Recent Media Coverage