Guess Who’s Back, Back Again…
Iowa Social Intelligence Data: Vivek Ramaswamy Data Outlier
***Update: READ our post below on the actual Iowa results***
**Original Post**
In 2020 we analyzed social intelligence data around the Presidential Election. Our work was featured in The Wall Street Journal, The New Yorker, Bloomberg, Yahoo, The Washington Examiner & many other outlets.
The goal was to learn how organizing and categorizing social data by battleground states could help explain narrative and trends going into Election Day. As I noted many times throughout the process, this was not meant to replace polling data, or be more/less accurate, it was meant to complimentary.
The “Presidential Social Intelligence Battleground Tracker” was an ongoing analysis using social intelligence and language comparison technology of how voters in battleground states discussed the candidates.
That project aggregated, analyzed, scored, and categorized well over 3.5 million social media posts and 65 million words into a Trump/Biden Supporter, Opposed, and Neutral Voter language data sets. The battleground states in our 2020 analysis were: AZ, CO, FL, IA, ME, MI, MN, NV, NH, NC, PA, TX, WI
Since 2020, social intelligence data has taken a larger and more defined role in the ongoing measurement mix in today’s politcal coverage — as I suggested could be the case in my 2020 Washington Times Oped:
FULL OPED: https://www.washingtontimes.com/news/2020/dec/8/whats-next-for-the-business-of-election-pollsters-/
Along the way we have continued our work using this data — mostly for our clients measuring public opinion and media impact. While we have been tracking data relative to the Iowa Caucus and New Hampshire Primary, we have not published anything publicly.
Ramaswamy Data Outlier
The reason for this analysis centers around an outlier we are seeing between social intelligence data and pubic polling data leading up the Iowa Caucus related to Vivek Ramaswamy.
In our past use of this data, one of the more interesting components has always been when there is a distinct outlier between public opinion, sentiment, and support index from social data vs. things like ballot test, and favorability in traditional polling.
With social intelligence data we don’t look so much for numbers/standing as a quantitative “moments-in-time” like with polling data — instead we look at qualitative/contextual trend-lines and how they compare to polling over time.
Our data is being aggregated via a custom measurement system in IA and NH from our technology partner EyesOver Technology.
— As of 1/13 Real Clear Politics has Vivek’s polling average in Iowa at 6.5%
— In contrast to the polling averages from Real Clear Politics above, the 7 Day Support Index in our EyesOver measurement system shows Vivek at 19% in Iowa, essentially tied for second with Ron DeSantis.
(Another interesting outlier is the diminished positioned of Niki Haley at 12% via the Support Index vs. her second place polling average in Real Clear Politics of 17.8%)
Support Index methodology can be found here
Again, it is important to note, we don’t look at this data and assume it is always correct/incorrect vs. differences in polling. But it is always worth at least noting when there are significant differences — especially when differences are sustained over time or part of a persistent trend-line leading to an Election Day.
In this case, the 30 day average (below) of the EyesOver Support Index shows Ramaswamy growing from 8% on December 13th to 20% as of January 12th, and interesting trend-line approaching Election Day.
This trend-line for Ramaswamy could be something to watch for on Caucus Night if the campaign is able to turn the support showing up for Ramaswamy in the online conversation into actual bodies at caucus locations.
Of course it is also a strong possibility that outsized vocal support across social media and digital channels (which this data relies on) could create an outlier in the opposite direction to polling data for a candidate like Ramaswamy (and inflating his numbers) in this type of analysis.
Discussion Volume & Sentiment
A corollary to The Support Index, and a factor that could prove to be a reason for the outlier in support for Ramaswamy via social data vs polling (or similarly an outlier undervaluing support for Haley using this type of data vs. traditional polling) is the volume of mentions and the sentiment around each candidate.
The data above is the 30 day average of mention volume and sentiment. Despite Niki Halley’s heavy media spending in Iowa you actually see Ramaswamy being mentioned more with higher sentiment consistently above 50%. Nikki Haley shows less overall volume and sentiment consistently below 50%.
**Again, conversation volume and sentiment do not always directly correlate proportionally to support at the polls — in fact there are many reasons why those metrics alone can differ significantly from polling and results.**
Large Language Model Analysis
The main aspect of social intelligence data that I have always loved is the massive language data set it provides for further analysis. Polling is very binary, but remains a trusted scientific measurement tool for good reason.
However, in today’s fast paced and over-messaged society, opinions can change quickly. And the nuance about “how and why” things change can sometime be discovered within the messy, self-expressed language in digital and social channels…if you can organize the clutter.
My favorite way to put it…the online conversation surrounding brands, issues of the day, and political figures is similar to one giant, real-time focus group. In this case, the data captured from the social conversations of Iowa voters can be processed and summarized via a Large Language Model.
**NOTE: TEXT BELOW IS AN AI GENERATED SUMMARY OF THE ONLINE LANGUAGE DATA SET FROM IOWA VOTERS AROUND THE ISSUES AND REASONS FOR THEIR SUPPORT OF VIVEK RAMASWAMY.***
Source: EyesOver
Vivek Ramaswamy
From the text provided, a few key topics emerge regarding the Republican Presidential Primary in Iowa.
1. **Vivek Ramaswamy’s engagement with constituents:** Several users mention the candidate’s active presence in the state and his willingness to interact directly with voters. They speak positively of his numerous stops in Iowa and his extensive engagement during these visits, suggesting that his hands-on approach is earning him support.
2. **Weathering the storm:** One user also commends Ramaswamy’s resilience, as he reportedly toured the state amidst a winter storm, contrasting his commitment with the perceived absence of other candidates like Nikki Haley and Ron DeSantis during similar circumstances. This demonstration of dedication and leadership is seen as a positive aspect.
3. **Support for Trump:** There’s also discussion about Ramaswamy’s alignment with former President Donald Trump. One user notes a conversation between Ramaswamy and a Trump supporter, where the candidate suggests that supporting him is the best way to “save” Trump. This association with Trump appears to be seen as a positive trait by some constituents.
4. **Comparison with other candidates:** Ramaswamy is contrasted with other potential candidates like Nikki Haley and Ron DeSantis. Some users suggest that these other candidates represent a part of the GOP that is “dying”, characterized by “endless wars, ineffective policy,” and a lack of transparency with journalists. One user also alleges that Haley’s campaign staff allowed them into an event, only to later accuse them of sneaking in. In comparison to these candidates, Ramaswamy seems to be gaining support due to his apparent authenticity and engagement.
You can read the full AI generated analysis here.
NOTE: TEXT ABOVE IS AN AI GENERATED SUMMARY OF THE ONLINE LANGUAGE DATA SET FROM IOWA VOTERS AROUND THE ISSUES AND REASONS OF THEIR SUPPORT FOR VIVEK RAMASWAMY.***
Final Questions
Things to watch on Monday in Iowa:
— Will the levels of support being measured for Ramaswamy via social data translate to support at the Caucus in similar proportions? Will he over perform polling?
— Will the lower numbers being measured for Nikki Haley translate to Caucus support or will she finish more inline with traditional polling?
— How will the weather impact turnout from various supporter bases?
Have fun out there everyone, good luck, be safe, and stay warm!
About The Author
Adam Meldrum is an award-winning political & media strategist. He is the Founder/President of the Republican media-buying operation AdVictory LLC. Adam has served as an advisor to other organizations in the political/technology space such as DDC Public Affairs, WinRed, & IMGE.
Adam also serves on The Board of Directors at HeadCount, a non-partisan organization that uses the power of music to register voters and promote participation in democracy.
Adam has been recognized as a 40 under 40 recipient by Crain’s Detroit Business, the American Association of Political Consultants, and a Rising Star by Campaigns and Elections.
Since 2006 Adam has produced award-winning work for and advised campaigns/organizations such as; Governor Rick Snyder, Governor Bruce Rauner, Governor Doug Burgum, Governor Bill Lee, Ambassador Ron Weiser, Senator Ben Sasse, Senator John McCain, Senator Rand Paul, The Republican National Committee and the U.S. Chamber of Commerce.
Adam’s work on the forefront of technology, data, and analytics was featured by the former OZY Media in a profile: “Meet the GOP’s Chatbot and Artificial Intelligence Guru”. He is a die-hard Michigan State University fan/alumni and pretty obsessed with Phish and The Grateful Dead. He resides in Washington D.C. & Birmingham, MI with his beautiful wife Christina, son Elvis, and their weird puppy, Basil.
Personal Website: http://www.adammeldrum.me/
Media Coverage
Our previous work around social intelligence analysis was previously featured numerous news outlets including The Wall Street Journal & The New Yorker:
Election 2020 Polls: Startups Pitch Themselves as Alternatives
Many campaigns, political groups still pay for polling, but use online surveys and social-media analysis to understand voters
The New Yorker
“Can We Trust the Polls?
Polls are not predictive; they are snapshots taken at a particular moment in time — one that will have passed by the time a poll reaches the public…”
New Hampshire Data
For context, at this point in time the data in New Hampshire between public polling and social intelligence support index looks as follows:
1/13 Real Clear Politics NH Polling Average:
1/13 EyesOver NH Support Index:
