Announcing: 2020 Social Intelligence Battleground Tracker
An ongoing measurement/analysis series examining social intelligence around the 2020 Presidential Election in battleground states
Battleground Project Overview
Social Intelligence:
This Medium publication leverages social intelligence, search trends, audience analytics, language analysis, and network graphs to analyze political campaign and news topics of the day.
Amongst other things, we examined the 2019/2020 Democratic Primary through Super Tuesday, Bloomberg’s impact, news coverage around the Kavanaugh hearings, COVID-19, a Democratic News Networks, and Mainstream Media Headlines over time.
For the rest of 2020, via a partnership with Eyesover Technologies, we will be analyzing social intelligence around the presidential election in battleground states.
Our analysis will be grouped by:
— Candidate & Topic Sentiment
— Candidate Support Gained and Lost by Day
— Candidate Head-to-Head Overall Support Index
Langauge Analysis:
In addition to scoring and measurement from this social intelligence data, we will also process a deep-dive into the language of Trump and Biden supporter data sets using Relative Insight’s language analysis platform.
The language analysis will:
— examine the frequency, similarities, and differences in the text used online to discuss the candidates in each state and by supporter classification (Trump Supporters, Biden Supporters, Neutral).
— compare the words, topics, phrases, grammar, and emotions in different slices of the overall data set.
— compare these data sets at different points in time, state by state etc…
States: AZ, CO, FL, IA, ME, MI, MN, NV, NH, NC, PA, TX, WI
Date Source: All data is publicly available and anonymized for our analysis from Twitter, Facebook, Online Blogs, & Message Boards.
Project Background
I began using social intelligence to measure the digital conversation around candidates and issues during my 2017-2018 consulting work on former Gov. Bruce Rauner’s reelection campaign. After 2018, I dove deeper via my corporate, government, non-profit, and trade association consulting work in the Omnicom network of companies at DDC Public Affairs.
Social intelligence technology in the political campaign and public affairs space has come a long way during this time.
In 2019, my firm, AdVictory, partnered with The U.S. Chamber of Commerce’s Political Program in a research initiative tied to their work in the NC-9 Special Election.
This work received four Reed Awards from Campaigns&Elections:
— Category, Overall: “Best Data Analytics Solution”
— Category, Overall: “Best Use of New Technology”
— Category, Research: “Best New Research Technique”
— Category, Research: “Best Use of Technology in Research”
In 2020, we have implemented this technology in research/analysis initiatives in many other client programs around the country.
Project Premise
The premise is simple — ongoing analysis/measurement of realtime conversations across digital channels helps expose insights and trends. I sometimes think of the internet as one, big, always-on, focus group. As social intelligence technology has evolved, it has sharpened its ability to weed out the garbage and process rich user-generated qualitative data into meaningful quantitative insights.
Big brands and corporations use this concept to measure how consumers feel about their products and/or services. It’s a natural fit for the political space.
So…away we go…
Qualifier
First, let’s start with our one major qualifier. This is not traditional polling.
In practical client implementations social intelligence pairs nicely with traditional research methods like polling, tracking, and focus groups. It does not replace it.
For the purposes of this work, it will stand on its own for analysis sake. But we will often compare our findings to public polling.
Social Intelligence: About Eyesover Technologies / Sample Data
About Eyesover:
“Eyesover provides real-time custom research by finding and analyzing online conversations.”
As mentioned above, on a state-by-state level:
Leveraging Eyesover’s Social Intelligence Technology, we will be examining:
— Candidate & Topic Sentiment
— Candidate Support Gained and Lost
— Candidate Overall Support Index
Overall Support Index:
This metric compiles over time from the whole of the digital conversation around each candidate (entity). It is based on measurement of postive/negative sentiment and support “gain/loss” scoring derived from how individuals signal their support for a candidate online.
Sample:
Topics and Sentiment:
As individual discuss each candidates, the conversation is grouped into political/policy/issue related topics based on keywords. Over time, we will be measuring the sentiment around each candidate in various topical categories.
Sample:
Support Gained/Loss:
One of the more interesting metrics we will be tracking in each state is an aggregate measurement of candidate support gained/loss by day. This data point is aggregated from individuals expressing support or opposition for a candidate in their online discussion.
It is classified into four tiers:
— Previous Non-Supporter to Neutral (Light Green)
— New Supporter (Dark Green)
— New Non-Supporter (Dark Red)
— Previous Supporter to Neutral (Light Red)
Sample:
Language Analysis: About Relative Insight Technology / Sample Data
“Relative Insight uses the power of comparison to highlight differences in the way specific audiences, and various demographics talk.”
Below are sample/example insights comparing language from Biden Supporters v. Trump supporters across every battle ground state from May 13th— June 18th. This data set comes from the anonymized social postings that makes up the daily “support gained/loss” data point.
Sample:
— 6,099,544 words from Trump Supporters
— 7,721,430 words from Biden Supporters
Battleground Biden supporters VS Trump Supporters:
— 2.1x relative difference in Topics/Words around “Death”
— 2.9x relative difference in Topics/Words around “No Caution”
— 44x relative difference in use of the phrase “president’s event”
— 127.5x relative difference in use of the phrase “come fall”
Battleground Trump Supporters VS Biden Supporters:
— 2.4x relative difference in Topics/Words around “Recorded Sound”
— 3.8x relative difference in Topics/Words around “Safe”
— 124x relative difference in use of the phrase “historical landslide”
— 169.5x relative difference in use of the phrase “enforcement arrest”
Sample:
Expanded Insights around Topics:
Summary
This summary post is meant to serve as background/overview for the overall project and as an example around the type of data we will examine through November’s General Election.
In practice, future posts will be much more specific to states, data sets, and points in time. Once a month we will present aggregate analysis.
About The Author
Adam Meldrum is an award-winning political & digital strategist. He is the Founder/President of the Republican media-buying operation AdVictory LLC, a Senior Advisor at DDC Public Affairs, and a Senior Advisor at WinRed. Adam 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.
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 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. with his beautiful wife Christina and their weird puppy, Basil.
Personal Website: http://www.adammeldrum.me/
